furthest voices in virgil’s dido i · furthest voices in virgil’s dido i ‘we have to stop...

FURTHEST VOICES IN VIRGIL’S DIDOI

‘We have to stop somewhere, but wealso have to face the fact that any par-ticular stopping-place is therefore ourchoice, and carries with it ideologicalimplications.’

Don Fowler 1

‘Magnus est Maro’.

I. L. La Cerda 2

Part one

Reading Dido – in the Aeneid and beyond – has always been anintensely charged literary and political game. A sensitive, lovingwoman, Dido offers Aeneas a real alternative to the complex businessof setting Rome in motion, and her death shows the enormous pricethere is to pay in terms of human fulfilment and happiness for the sakeof empire building. How more or less sympathetic and straightforwardshe is seen to be is of course crucial to our perception of Aeneas as epichero, and to the meaning of the Aeneid as a whole. It is only naturalthat throughout the 20th century, and into the 21st, critics have over-whelmingly packaged this fascinating character as the archetypical‘other voice’ to the poem’s teleological (not to say ‘Augustan’) plot.

1 D. P. FOWLER 1997: 25 = 2000: 127-128. 2 LA CERDA: vol. 1, p. 441.

Furthest Voices in Virgil’s Dido

This comfortable opposition rests to a significant extent on a one-sided reading of Dido’s emotional intricacies. Already ancient poets andreaders, from Ovid to the Christians, vigorously edited Virgil’s Dido toproduce their own challenge to his epic. Building upon Heinze’s influ-ential treatment, modern critics have favoured a comparable approach:current readings emphasize the image of a loving and forlorn heroinewhose short outbursts of rage and fury are evanescent – and justified.Yet as I set out to show here, a reading that is willing to delve furtherinto the intertextual complexities of Dido’s character, and to probe therelation between intertextual trace and unconscious impulse 3 inBook 4, reveals a darker, more thrilling and suspenseful narrative thanmodern critics have often allowed themselves to imagine.

This paper started life as a note on a famous and much debatedpassage at the heart of the book, where Dido’s intricate, opaque lan-guage appears to convey more than meets the eye. I decided to retainits original inductive structure even as it grew to chart more widely thecomplexities of Virgil’s construction of Dido’s character in other sec-tions of Book 4, and Ovid’s rewriting of Dido in his Heroides 4.

1. At a crucial juncture in the sequence of events that lead toAeneas’ departure and Dido’s suicide, the queen makes a desperateattempt to persuade Aeneas to postpone his plans. She addresses hersister Anna and begs her to approach him (424-436) 5.

3 I hope to be able to offer a theoretical justification of my reading strategy else-where. Here I will freely resort to concepts such as ‘textual unconscious’ (with impliedinverted commas) trusting that the way I use them is clear enough, but I also want toacknowledge an important debt to Francesco Orlando’s masterful work. It is in anycase worth pointing out that whenever I attribute feelings and emotions (conscious orotherwise) to Dido or other characters I actually intend to refer to the text’s own lin-guistic construction of those feelings.

4 This paper is part of a wider study of Dido’s characterization and models. Itshould be read in conjunction with SCHIESARO 2005, which develops some closely con-nected issues.

5 Text and apparatus, here and elsewhere, are MYNORS’. Commentaries (and basicworks of reference) are referred to by name only, except when the date of publicationis relevant to the discussion and is therefore specified, or when reference is made tothe introduction or other narrative parts. A convenient list of Virgilian commentariescan be found in KNAUER 1964: 13-17.

61

Alessandro Schiesaro62

‘i, soror, atque hostem supplex adfare superbum:non ego cum Danais Troianam exscindere gentem 425Aulide iuravi classemve ad Pergama misi,nec patris Anchisae cineres manisve revelli:cur mea dicta negat duras demittere in auris?quo ruit? extremum hoc miserae det munus amanti:expectet facilemque fugam ventosque ferentis. 430non iam coniugium antiquum, quod prodidit, oro,nec pulchro ut Latio careat regnumque relinquat:tempus inane peto, requiem spatiumque furori,dum me mea victam doceat fortuna dolere.extremam hanc oro veniam (miserere sororis), 435quam mihi cum dederit cumulatam morte remittam.’

436 dederit MPp : dederis w Seru. cumulatam Ppw, Seru. : cumulataMb?; -ris -tam ‘Tucca et Varius’, -rit -ta ‘male quidam’ ap. Seru.

Already DServius 6 reached the conclusion that Dido here perplexelocuta est 7, and in fact the entire speech is dense, allusive, and com-plex, as befits a tragic heroine who is gradually realising the enormityof the approaching catastrophe. The peculiarly intricate last few lineshave much worried scholars – again, at least since Servius. Just aboutevery word in line 436 has spurred sustained critical analysis: the alter-native between dederit and dederis; between cumulatam and cumulata;the interpretation of morte and that of remittam, have all been hotlydebated. In spite of Peerlkamp’s despondent prediction that ‘haecnemo unquam intellexit, neque intelliget,’ a reasonable degree of cri-tical consensus appears to have crystallized around the text printedabove since the middle of the 19th century, with Ribbeck’s critical edi-

6 DServius on 436, quoted in full later, p. 74. Cf. PÖSCHL’S 1962: 84-85 succinct butuseful analysis.

7 Among later commentators, LA CERDA (apart from displaying his usual mastery ofliterary models and nuances) is more attuned than most to the complexities and ambi-guities of Dido’s words, in Book 4 in general, and in this section in particular – termssuch as dissimulare, insinuare, latere, ambages, circuitus recur. See for instance his sharpnote on 419-420 (LA CERDA: 441), where he rightly argues that si potui tantum speraredolorem, / et perferre, soror, potero ‘insinuat, non posse se perferri, quia id non potuitsperare’, or on 422 te colere (cf. n. 172). I plan to return to the issue elsewhere.


tion (1859) 8. Pease’s paraphrase gives a view shared by most contempo-rary interpreters: ‘I beg this last favour – pity your (poor) sister –, andwhen he shall have granted it my death will return it with interest’ 9.Aeneas, not Anna, should grant the venia of a short delay, theextremum … munus mentioned at 429: hence dederit, not dederis, a sim-plification predictably inspired by miserere sororis in the line before 10.Pease intentionally preserves almost all the different nuances enshrinedin the ablative morte 11. Cumulatam refers back to veniam 12, in a wellestablished use of the verb (‘to increase, augment, enhance’) 13, whileremittere takes the place of the more common referre or reddere 14.

8 The text dederit cumulatam is then printed by LADEWIG, SCHAPER6 (1870); FORBIGER4

(1872-75); SABBADINI (1884-881); PAGE (1894-1900); MACKAIL (1930); SABBADINI (1930); BU-SCAROLI (1932); PEASE (1935); PARATORE (1947); AUSTIN (1955); MYNORS (1969); GEYMONAT

(1973, 2008); WILLIAMS (1972). On dederis cf. n. 10, on cumulata below pp. 75-81. 9 PEASE on 436. Cf. also AUSTIN on 436: ‘and when he has granted this kindness to

me, I shall repay it a thousandf old at my death.’ 10 Dederis was long the standard reading, accepted among others by LA CERDA,

HEINSIUS, BURMAN, HEYNE-WAGNER, up to and including FORBIGER3 (1852). After the successof RIBBECK’S choice of dederit, only CONINGTON, NETTLESHIP retain dederis. J. HENRY 1853and 1857 prefer dederit, but J. HENRY 1873-92 returns to dederis. VIVONA 1898: 431believes that the bulk of 416-436 was originally spoken by Dido herself to Aeneas, butVirgil later changed the narrative pattern for variation’s sake: dederis, which Tucca andVarius must have read in Virgil’s autograph, would testify to the original plan. On thecontrary, SABBADINI 1900 argues that the original version of the passage (without 431-434, and with no mention of Anna), featured dederit, with morte as an instrumental;subsequently Virgil introduced Anna, switched to dederis, and gave morte a temporalsense. But SABBADINI’S 1930 edition prefers dederit, cf. also SABBADINI 1910-20: IX-X, wheredederis is regarded as an ‘espediente provvisorio.’ This primitive version of the textlacked 431-434, and assigned no role to Anna; but in the later version cumulatam morteremittam contradicts the context – and dederis was therefore revived by those seekingto eliminate the contradiction (against the possibility that surviving mss. of the Aeneidmay preserve variant readings see in general GEYMONAT 1995: 297-298). A recentattempt to revive dederis (MURGIA 1987) has been effectively refuted by CASALI 1999.

11 The addition of ‘my,’ however, is not neutral, and occludes possible furthermeanings: see paragraph 3 below.

12 Contemporary editors, unlike most of their predecessors, usually avoid placing acomma after dederit. Some ‘old editions’ punctuated rather after cumulatam (CONINGTON

on 435-436). 13 OLD s. v. 5, e.g. Cic. Phil.14.30 ea [sc. praemia] quae promisimus studiose cumulata

reddemus.14 Cf. Liv. 24.48.3 ut in tempore et bene cumulatam gratiam referant, but cf. Mart.

5.59.3 quisquis magna dedit, voluit sibi magna remitti.

63


I will return to these and other issues later, for in the first instanceI want to dwell on a point which does not appear to have attractedmuch curiosity, let alone concern, namely the nature and import of thevenia which Dido begs of Aeneas 15. Her request appears simpleenough: tempus inane peto, requiem spatiumque furori (433). Themeaning of inanis is sometimes difficult to pin down in Virgil 16, but inthis case commentators and translators appear to agree: ‘a few vacanthours’ (Conington); ‘time without action (on your part)’ (Pease); ‘amere nothing – just time’ (Day Lewis) 17; ‘time, a blank time’(Austin) 18. At one level inane is meant to appeal to Aeneas’ point ofview – the time Dido asks for would be of little use to him anywaybecause weather conditions are not favourable to navigation in thewinter season: he might as well await favourable winds and an easydeparture, facilemque fugam ventosque ferentis (430). But inane alsoreassures Aeneas that, should she be granted it, Dido will not use thistime to try and rebuild their relationship or to persuade him to stay atCarthage, as she makes clear at 431-432 and as Servius clumsily trans-lates in the promise that she would not seek sexual intimacy (sine offi-cio coeundi) 19. It is not essential, at this stage, to enquire whether Didois consciously dissimulating her hopes to revive the coniugium antiqu-um or is actually convinced that all hopes are to be abandoned, and isready to resign herself to a final farewell. The fact, however, that she isdeploying an argument Anna had already suggested as a way to delay

15 A different kind of veniam had already played a role earlier in the book, atAnna’s own suggestion: tu modo posce deos veniam, sacrisque litatis / indulge hospitio (50-51). The repetition at 436 is part of a pointed series of contrasts between the begin-ning of Dido’s love story and its tragic conclusion, cf. p. 105.

16 See SCARCIA 1984-1991. In this context there may be a shade of the technicalmeaning inane tempus = keno;" crovno" = ‘blank time’ (OLD s. v. 8) as used in ancient dis-cussions about music and metrics, e.g. Quint. 9.4.51: cf. M. L. WEST 1992: 266 nn. 30-31, and, below, n. 38.

17 Cf. ‘mere time’ (CLAUSEN 2002: 93), all in J. HENRY’S footsteps (1873-92: 3.175). 18 ‘Poco tempo’ (CANALI), ‘un momento di requie’ (SCARCIA), ‘tempo solo’ (CALZECCHI

ONESTI), ‘un moment, presque rien’ (PERRET; cf. PEERLKAMP: ‘quasi rem non magnam’). 19 Cf. LA CERDA (explicatio ad 425-436): ‘[sc. inane] nam illud erit exigui commodi,

cum caritura sim voluptate matrimonij’, and TILLY’S decorous translation (London 1968):‘free from passion’.


Aeneas’ departure and thus creates a chance for their relationship toblossom is in itself revealing 20.

No intertextual models are usually invoked for Dido’s request, yeta remarkably similar one plays a crucial role in a text to which thisbook of the Aeneid is indebted in many important respects. In Euripi-des, Creon orders Medea to go into exile wJ" tavcista (321) 21, but thewoman reassures him of her good intentions and obedience, and begsfor a short reprieve (340-342):

mivan me mei`nai thvnd∆ e[ason hJmevrankai; xumperanai frontivd∆ h|i feuxouvmeqapaisivn t∆ ajformh;n toi`" ejmoi`" ...

Creon reluctantly agrees (350-356) 22:

kai; nu`n oJrw` me;n ejxamartavnwn, guvnai,o{mw" de; teuvxhi tou`de. prounnevpw dev soi,ei[ s∆ hJ ∆piou`sa lampa;" o[yetai qeou`kai; pai`da" ejnto;" th`sde termovnwn cqonov",qanhi: levlektai mu`qo" ajyeudh;" o{de.nu`n d∆, eij mevnein dei, mivmn∆ ejf∆ hJmevran mivan:ouj gavr ti dravsei" deino;n w|n fovbo" m∆ e[cei.

Her request to Creon distinguishes Euripides’ Medea both fromher Apollonian counterpart and from some related characters such asHypsipyle and Ariadne, and is essential to the development of her plotof revenge: it is precisely thanks to the short delay extracted from theking – the one day which dramatic conventions require for the unfold-ing of a tragic plot – that Medea can accomplish her plans, as sheremarks in the monologue which follows (371-375):

oJ d∆ ej" tosou`ton mwriva" ajfivketow{st∆, ejxo;n aujtwi ta[m∆ eJlein bouleuvmata

20 4.51-53 indulge hospitio causasque innecte morandi, / dum pelago desaevit hiems etaquosus Orion, / quassataeque rates, dum non tractabile caelum. Ovid will expand thepoint, see later p. 215 (SIFC 2/2008).

21 Cf. also 274, 335. 22 Lines 355-356 are excised by NAUCK, followed by several editors (a full list in VAN

LOOY 1992: 126), including, recently, DIGGLE, but not PAGE, VAN LOOY (with reservations)and MASTRONARDE. A few editors condemn 356 but spare 355.

65


gh" ejkbalovnti, thvnd∆ ejfh`ken hJmevranmei`nai m∆, ejn h|i trei" tw`n ejmw`n ejcqrw`n nekrou;"qhvsw 23, patevra te kai; kovrhn povsin t∆ejmovn.

Indeed a brief delay is all she needs: meivnas∆ ou\n e[ti smikro;ncrovnon, É ... É dovlw mevteimi tovnde kai; sigh`i fovnon (389-391) 24.

Virgil’s tempus inane is consistent with mivan ... hJmevran and smikro;ncrovnon (note for instance Conington’s translation), but a further impli-cation emerges through the intertextual model: inane conveys, with adifferent emphasis, the same reassurance implicit in Medea’s insistenceon mivan, the single most important element of her exchange withCreon. Thanks to her insistence, Creon himself is convinced that theshort time he finally grants Medea will not be enough for her to commitany of the crimes he fears. In the light of this model, however, Dido’sinane rather rings as an unrequested apology, almost a Verneigung:Medea’s time has surely been anything but ‘empty’ 25. Is it reasonable tosuspect that Dido’s request, too, is fraught with similar dangers?

Seneca in his Medea will make Creon’s train of thought even moretransparent (285-296) 26:

ME. Per ego auspicatos regii thalami toros,per spes futuras perque regnorum status,Fortuna varia dubia quos agitat vice,precor, brevem largire fugienti moram,dum extrema natis mater infigo oscula,fortasse moriens. CR. Fraudibus tempus petis.ME. Quae fraus timeri tempore exiguo potest?

23 Here Medea appears to be planning Jason’s death as well (MASTRONARDE ad loc.),but if the future tense is seen as modal then a suggestio falsi rather than an outrightfalse prediction may be involved (COLLARD 1988: 314 n. 10).

24 An alternative version of the Carthaginian myth of origins (Suda s. v. Foinivkwnejgkevfaloi) also involved the request for ‘a little time.’ Stranded on the coast of Libya,the Phoenicians refugees ask the locals to welcome them nuvkta kai; hJmevran, which thelatter interpret as ‘just one day and one night’, the former as ‘night and day’, i.e. ‘forev-er’: SCHEID, SVENBRO 1985: 336-337 (the ruse recurs in other Greek myths of foundationas well). Cf. n. 93 below.

25 HEYNE on 433: ‘tempus inane est, quo nil agitur’ (quoting Val. Fl. 3.656 has toleraremoras et inania tempora).

26 For Ovid’s reworking of the Virgilian scene in his Heroides see later, p. 214 (SIFC2/2008).


CR. Nullum ad nocendum tempus angustum est malis.ME. Parumne miserae temporis lacrimis negas?CR. Etsi repugnat precibus infixus timor,unus parando dabitur exilio dies.ME. Nimis est, recidas aliquid ex isto licet;et ipsa propero 27.

Indeed, Seneca amplifies the resonance of the dies-motif by havingMedea reflect explicitly – twice – on its strategic importance, first at399-400 (segnis hic ibit dies, / tanto petitu ambitu, tanto datus?), thenmore extensively at 420-425:

laxare certe tempus immitis fugaegenero licebat – liberis unus diesdatus est duobus. non queror tempus breve:multum patebit. faciet hic faciet diesquod nullus umquam taceat – invadam deoset cuncta quatiam.

As he rewrites the tragedy of Medea in the form of a Virgilian centoheavily influenced by Seneca’s play, Hosidius Geta will use Dido’swords to express Medea’s request for extra time (78-80): non iam con-iugium antiquum, quod prodidit, oro: / tempus inane peto: liceat subdu-cere classem; / extremam hanc oro veniam.

Ennius’ Medea exul 28 probably contained a similar exchangebetween Medea and Creon. Although we cannot read the whole scene,lines 229-231 Jocelyn hark back to Medea’s considerations at Eur.Medea 373-375, and show that this Medea, too, is fully aware that theday’s delay foolishly accorded her is a formidable weapon for her ira 29:

27 Medea’s nimis est is also double-edged. She ostensibly reassures Creon that aday will be plenty for her to prepare her exile, and she is as eager to leave at thisstage as he is too see her leave. But nimis est, taken in isolation and set off by thepause which follows, also means what it says – the time Creon has unguardedly grant-ed her will undoubtedly be ‘too much’.

28 On the title of the play and the attribution of fragments see JOCELYN 1967: 342-350, especially 346, 348.

29 See JOCELYN 1967 ad loc. A detailed stylistic analysis of the fragment in LENNARTZ

(1994) 212-238, with further bibliography.

67


ille traversa mente mi hodie tradidit repagulaquibus ego iram omnem recludam atque illi perniciem dabomihi maerores, illi luctum, exitium illi, exilium mihi.

2. Dido’s request for a short reprieve has attracted almost universalsympathy, and critics have therefore proved unusually keen to take herwords at face value, at least in this respect. All she asks for, after all, isa bit of ‘empty time’ with no obligations and no consequences forAeneas’ mission: vis à vis such an innocent demand his refusal appearseven more hard-hearted and callous 30.

The sympathetic interpretation of tempus inane has encouragedcritics to smooth out considerable difficulties in the analysis of Dido’sspeech even before we reach its tormented last line – famously ‘themost difficult in Virgil’ (Conington) 31, or at least ‘an outstandingexample of Virgilian mystery’ (Austin) 32.

Let us consider, first, requiem spatiumque furori, which rounds off,probably more as a predicate than an apposition, line 433 (‘someempty time, which would be requies et spatium for my furor’). ‘Spatiumalone would be difficult with furori, but requiem spatiumque combineto mean ‘a resting-time’, Austin shrewdly observes: he is not alone inacknowledging, then silencing, the difficulty. Requiem spatiumque maywell be modelled on the more common tempus et spatium, but thelatter does not contain any potential dissonance, because tempus andspatium neatly overlap (almost redundantly), whereas requies andspatium – effect and cause – do not. There is no ambiguity, forinstance, in the nurse’s recommendation to Clytemnestra, in Seneca’s

30 In the economy of the plot as engineered by Juno and Venus at the beginning ofthe book (90-128) time is not altogether irrelevant; as Anna recommends in her firstaddress to her sister she should create causas … morandi (51) to detain Aeneas.

31 A definition echoed in the title of SABBADINI 1900: ‘Il verso più difficile dell’Eneide(IV 436)’.

32 So difficult and mysterious, indeed, that PEERLKAMP and GOSSRAU could not counte-nance Virgilian authorship, and excised it alongside 435 (LA CERDA, explicatio ad 425-436, mentions earlier opinions that miserere sororis was a gloss, to be deleted togetherwith 436). WAGNER, too, wonders (ad loc.) whether 436 should not be excised. Cf. BEL-LESSORT 1925: 115: ‘Ce passage est probablement un de ceux que Virgile eût modifié s’ilavait pu corriger son poème.’


Agamemnon, to take time before heeding her rage: proin quidquid est,da tempus ac spatium tibi: / quod ratio non quit saepe sanavit mora(129-130), even if, interestingly, the queen seems far from convincedby this line of thought: tragic characters seldom live their lives accord-ing to Otto’s Sprichwörter 33.

But in Virgil’s line, as Paratore points out (ad loc.), the Latin wordscan have the desired sense of ‘a resting-time’ only if we allow that bothhysteron proteron and zeugma are simultaneously at work: logicallyspatium must precede its result requiem, but, unlike requiem … furori,spatium … furori unlocks a rich seam of exegetical ambiguity. Transla-tors and commentators generally iron out the problem by tacitly imply-ing that the one correct interpretation of spatium is that of ‘interval’ 34:‘a respite and breathing-space for my passion’ (Conington); ‘requie eintervallo al furore’ (Canali); ‘a breathing space for her passion’(Clausen); ‘rest and respite for my passion’ (Austin), ‘pausa e pace alfurore’ (Calzecchi Onesti). Only a few strive to preserve the complexityof the Latin word-order – notably Perret: ‘un peu de calme, quelquesjours pour mon délire’ 35. Ovid allows for no ambiguities: tempora parvapeto, / dum freta mitescunt et amor, dum tempore et usu / fortiter ediscotristia posse pati (Her. 7.178-180) 36. Yet it is well-known that Virgildoes exploit the ‘momentary and evanescent’ ambiguity produced byjuxtaposing words which, once the full meaning of the sentence is clear,turn out to be otherwise syntactically connected 37.

33 For further comparisons with the Seneca passage see TARRANT ad loc. and OTTO

1890 s. v. mora. St. Th. 10.703-705 harks back to Virgil’s line: ne frena animo permittecalenti, / da spatium tenuemque moram, male cuncta ministrat / impetus.

34 This meaning of spatium is rarer (OLD s. v. 9), and normally used with verbswhich signal it.

35 It may be interesting to note that MACKAIL (‘the last of Virgil’s great Romantic crit-ics’: QUINN 1963: 31) and DAY LEWIS, who appear to be at least partially attuned to thepotentially ominous implications of these lines, are both poets rather than, or morethan, professional classicists. On MACKAIL (1859-1945), who trained as a classicist butbecame Professor of Poetry at Oxford in 1906, cf. also S. J. HARRISON 1990: 3.

36 On the meaning of Virgil’s doceat dolere and its difference vis à vis Ovid’s ediscopati see later, p. 215 (SIFC 2/2008).

37 CLAUSEN 2002 (19871): 43 (from which my quote is taken), 71-72, 77, cf. O’HARA

1997: 250-251.

69


The problem is real. Assuming that spatium here means ‘(a stretchof) time’ – as it normally does when followed by a dative – 38 a literaltranslation of the iunctura would have to be ‘peace[,] and time forfuror’, a distinctly ambiguous and potentially disturbing meaning espe-cially if we consider the pause after requiem 39. We should not hastento paper over this difficulty by ‘decoding’ the Latin words, but, rather,recognise the rewarding interpretive scenario opened up thanks to thedouble-edged meaning of spatium furori alongside requiem furori 40.Again, it will not do to force an optimistic resolution by implyingleniendo or the like, as, for instance, Day Lewis does as he translates:‘…just time to give rein to despair and thus calm it’ 41. Spatium furorimay also be interpreted as ‘time for my furor to calm down’, becauseancient ethical theory (and common-sense advice) do point out thatthe best way to deflate anger and other strong emotions is to allowthem to run their course for a while (obstacles can only intensify theirstrength, as Ovid points out in the case of Pentheus: Met. 3.566-71);but this is neither the only possible interpretation of Virgil’s line, norindeed the most straightforward.

A theoretical restatement of this type of advice is offered by Senecaat de ira 3.39. At first, ira cannot be contained: primam iram non aude-bimus oratione mulcere: surda est et amens. Dabimus illi spatium (39.2);afterwards, quies will help to assuage anger (39.3 initia morborumquies curat): omni arte requiem furori dabit (39.4). Both Clytemnestra’sreaction to the nurse’s advice, and Seneca’s reasoning in de ira makeclear that ‘giving time to furor’ retains the more ominous sense of ‘atime and space for my fury’: things will definitely get worse before they

38 OLD s. v. 10. Note, in connection with the use of inane discussed above, n. 16,that spatium is also used to denote ‘a pause’ in music (OLD s. v. 16).

39 Cf. for instance Valerius Flaccus 2.356: et deus ipse moras spatiumque indulgetamori. The deus is Jupiter, the lovers Jason and Hypsipile.

40 While it is natural to assume that Dido is talking here about her furor, her wordsmay also be taken, in part, as a criticism of Aeneas’ own furor, the cruel folly (see e.g.4.309-311), as she sees it, of his hasty departure.

41 An early suggestion in LA CERDA (explicatio on 425-436): ‘peto requiem spatiumque,non amoribus, non blanditiis, sed furori meo mitigando … Solet enim primo impetusesse furor, qui postea assuetudine est dolor’.


can get better, before ratio will eventually be in a position to reassertits control over passion. Spatium is the first step in the path to recov-ery, but it is also a time for anger to express itself without restraint,because there is no reasonable way to contain such an outburst of pas-sion in its prime. But what can happen while furor is allowed spaceand time to spend itself?

3. Arguments about the meaning and implications of the ablativemorte 42 at line 436 also date at least as far back as Servius 43. Threeclosely connected issues are at stake: (i) the precise syntactical functionof morte (instrumental, temporal, or both?) 44; (ii) which type of death

42 There is no need to emend morte, and attempts have in any case been discour-aging. The least perverse proposal is probably the Delphin edition’s cumulatum (‘I willsend him away with my death to crown and reward him’ [Conington]), whose aggres-sive sarcasm is quite extreme (hence Klouacek’s cumulatum munere mittam). Ribbeck,not on form, proposed monte (RIBBECK 1866: 95: ‘[i]immo magnos conieci reginampromisisse montes auri argentique aliorumque praemiorum in tam exigui beneficii gra-tiam’); Schrader ventured sorte, which is too technical. VIVONA 1898: 430-431 arguesthat Dido must have expressed the same feelings of her Ovidian counterpart at Her.7.181, something along the lines of ‘s’ei mi concederà questo favore, mi salverà dallamorte’, i.e. (tentatively) quam mihi cum dederit, tum ablata morte remittam. Otherattempts are listed – and rejected – by PEASE 1935: 361. HEINZE 1915: 115 n. 38 inter-prets the transmitted text thus: ‘I will reward it in good measure, even with my life,’while remarking that the meaning of Dido’s ‘obscure’ words would have been clearerif the author had written cumulatam vel morte remittam, i.e. ‘even in death.’

43 Morte plays an equally interesting role at 4.17 postquam primus amor deceptammorte fefellit, where LYNE 1989: 31 brilliantly shows that the ablative is at work withboth fefellit and deceptam. The equation between death and deception, in itself plausi-ble given the context, can also be seen as an early instance of a propensity to amalga-mate heterogeneous categories which later grows to more alarming proportions, seelater p. 206 (SIFC 2/2008). On similar ambiguities created by ablatives in Virgil seeJACKSON KNIGHT 1966: 264-265 and O’HARA 1997: 249-250.

44 LA CERDA (explicatio ad 425-436), who reads dederis, recognizes that morte couldmean either ‘cum ipsa moriar’ or ‘si opus fuerit, moriar pro te’, but does not suspectthat suicide may be hinted at (though Wagner may have him in mind among thosewho ‘[d]e morte voluntaria … haec accipiunt’, and with whom he disagrees ad loc.).FORBIGER2 (1845-46) on 436 first explicitly remarks that morte could equally well beseen to refer to Dido’s natural death, or to her suicide: ‘Annam quidem haec de natu-rali morte intelligere vult soror, ipsa autem iam de caede sibi inferenda cogitat.’ Theambiguity between the instrumental and temporal meanings is recognized by severalmodern commentators (CONINGTON, PAGE, PEASE, BUSCAROLI, WILLIAMS), but denied byAUSTIN. SABBADINI 1900 argues that the meaning shifted from the instrumental to the

71


is alluded to (natural or voluntary?); and (iii) who is expected to die(Dido alone, or others as well?).

Word order, context and Latinity make it natural to consider morteas an instrumental. ‘By her death’, Dido would align herself to unluckylovers such as Damon in Eclogue 8 45, who, when he is spurned byNysa, presents his death 46 as an extremum … munus (Ecl. 8.60) 47. Theconnection between Dido’s and Damon’s words is clearly flagged byextremum hoc miserae det munus amanti at line 429 48. Her deathwould be a ‘gift’ to Aeneas, presumably because it would remove anyremaining obstacle to his plans 49. If this passive-aggressive threat wereunequivocally perceptible, however, Anna would have cause to worryabout Dido’s suicidal plans. But with a reasonable amount of goodwillmorte could be regarded as a less threatening indication of time 50:

temporal in the context of Virgil’s reshaping of the passage (see above n. 10). PARATORE

(on 436) and CARTAULT 1926: 324 privilege the instrumental meaning alone, but high-light (esp. Paratore) the ambiguity of the line. Recent discussions in MURGIA 1987: 56-57and esp. CASALI 1999: 106-113.

45 As Servius (on 436) already pointed out. 46 The connection between munus and suicide is debated, see CLAUSEN on Ecl. 8.60.

The model is Theocritus, Id. 3.25-27 (with HUNTER ad loc.); cf. [Th.] Id. 23.20-21. 47 The dialogue with Eclogue 8 is extensive. Shortly before his passive-aggressive

threat, Damon devotes a stanza to Medea (the passage is complex and much dis-cussed, see COLEMAN and CLAUSEN ad loc.): saevus Amor docuit natorum sanguine matrem /commaculare manus. crudelis tu quoque, mater. / crudelis mater magis, an puer improbus ille?(47-50). Saevus Amor harks back to Ennius: Medea animo aegro amore saevo saucia(Medea exul 216 Jocelyn), which also stands behind the all-important opening line ofBook 4 (mediated, there, via Cat. 64.249-250, but also looking at Apollonius’ Medea,3.286-287; see CLAUSEN 2002: 75-76, and later, n. 327), while improbus is echoed at Aen.4.412 (see later, p. 92 and n. 146). Also, Ecl. 8.64-109 describes a magic incantationclosely mirrored in Aen. 4.506-521 (FARAONE 1989; SCARCIA 1991): see later.

48 A comparable train of thought is found in Apollonius’ Medea, who would preferto be slaughtered by Jason rather than be abandoned by him: her death – she pointsout bitterly – would come as ‘due gratitude’ (4.375 ejpivhra ... ejoikovta) for all the helpand love she has bestowed on him.

49 Servius’ matter-of-fact reasoning –‘nam si eam odio habet restat ut eius mortelaetetur’ – resonates in Dryden’s harsh translation (with dederis): ‘If you in pity grantthis one request, / My death shall glut the hatred of his breast.’ TRAGLIA 1983: 153 para-phrases ‘mi conceda quest’ultimo favore, che gli restituirò ridandogli piena libertà conla mia morte.’ A similar interpretation in FAIRCLOUGH, BROWN 1908 ad loc.: Dido would beplanning to ‘deliver him [sc. Aeneas] from her forever.’

50 A clear-cut temporal use of morte without an adjective or genitive is hard to paral-lel even in archaic authors. Already Plautus favours in morte (LEUMANN, HOFMANN, SZANTYR


Dido would simply be promising that ‘at her death’ (at whatever time,and for whatever cause) she will gratefully reward Aeneas’ venia 51.These radically different messages have different addressees in sight.Anna, not for the first or the last time credited with a limited flair forlateral thinking, is not expected to surmise that ‘at death’ may point toan imminent future, let alone to pick up the Eclogue intertext. At somelevel, however, Dido must expect Aeneas to detect the desperate mes-sage lurking in her words, and – hopefully – be persuaded to reconsid-er his departure.

The absence of any explicit indication of whose death is meant withmorte, however, gives Dido’s words a threatening overtone consonantwith the implication of the Euripidean intertext we have been dis-cussing – and this is far from surprising at the conclusion of a speechwhose very first line saw Aeneas metamorphosed from a cherishedhospes (4.10) 52 into a haughty hostis 53. DServius (on 436) reports that

147-148); at Miles 707 mea bona mea morti cognatis didam (LINDSAY), where in any casemea qualifies morte, LEO adopts LINDEMANN’S emendation mea bona in morte, quoting asparallels Men. 411 and Cic. Caec. 10. Cf. also BENNETT 1914: 381. But occasionally mortecan be regarded as a borderline case between causal and temporal. LEUMANN, HOFMANN,SZANTYR 132 quote Lucil. 755-756 M. morte huic transmisit suam / scolen, to which onecould add Ter. Andria 799 eius morte ea ad me lege redierunt bona, and Virg. Aen. 3.333-334 morte Neoptolemi regnorum reddita cessit / pars Heleno (HORSFALL ad loc.: abl. eithercausal […] or perhaps of time, as used by V. with words not themselves of any tempo-ral force; cf. 629 [discrimine tanto], 1, 672 tanto cardine rerum, 4, 502 aut graviora timetquam morte Sychaei, with ref. to ANTOINE 1882: 189 ss., 232 ss.).

51 There is no need, and no way, to pin down what the queen may specificallyhave in mind as a reward. MURGIA 1987: 56, reading dederis, thinks, along establishedlines, of ‘an inheritance, and probably succession as queen.’ CASALI 1999: 111-112 and118, with dederit, believes that Dido is offering a retraction of her curse (cf. 4.380-384),more favourable travel conditions, and, by extension, a future of peace betweenRomans and Carthaginians.

52 On Aeneas as hospes see GIBSON 1999. 53 The emphatic hostis at 424 could well be gesturing towards Euripides’ Medea and

its pervasive use of ejcqrov". Medea consistently regards Jason and his family as herenemies (278, 383, 734, 750, 765, 767, [782], 797, 1050, 1060; note the concentration inMedea’s speech at 764-810), whom she must conquer (374, 765, 921). Medea sees her-self as an equal to Jason, who by his betrayal has violated the social norms regulatingrelations among xevnoi: ‘Medea thus takes on the traits of the insulted chieftain’ (MA-STRONARDE 2002: 9 and on 383). Jason is indeed e[cqisto" (467; he will return the com-pliment at 1323). Aegeus and the Nurse share her point of view as well as her

73


the potentially aggressive meaning of morte taken as an instrumentalhad been spotted by early critics: ‘alii ita intellegunt: ‘reddam illi grati-am, occidam illum’; nam alibi ait (600) non potui abreptum divellerecorpus? sed hoc totum sorori dicit.’ Very few modern interpretersacknowledge that ‘by death’ at line 436 could also be read as a threat toAeneas 54, and those who do hasten to rule out the possibility, perhapsbecause they implicitly share Mackail’s warning that ‘it would be quiteout of place to make [Dido] end this piteous appeal on a note of savageirony’ 55. Not DServius, however, who precisely at this stage in hisexplanation adds the intriguing remark ‘an perplexe locuta est, utsolent loqui mali aliquid molientes?’ (my emphasis) 56.

This is plainly a case in which the decision of where to stop in thesearch for ‘further voices’ is shaped by our exegetical expectations 57.If Dido is to be the forlorn victim of Aeneas’ callousness, she cannot atthis stage, as she begs for a short delay, be allowed to nurture any vio-lent thoughts against the man she is still desperately trying to hold onto. But Austin misses the point when he thinks that expanding thereach of morte to include Aeneas as a victim would represent Mackail’s‘note of savage irony.’ First of all, this possible interpretation shouldnot be seen as an alternative, but as an addition to the passive-aggres-sive meaning of morte which would foreshadow Dido’s suicide. Se-condly – and this is the essential point – there is no question of Didobeing conscious of the threat against Aeneas which, try as we might,

vocabulary (95, 744). The significance of the motif is well explained by Medea herselfat 506-508 – Jason’s actions have caused an inversion of normal affective relations, andas a consequence she has become ‘an enemy to my own kin … and made foes ofthose I ought not to have armed.’ The Nurse had analysed the situation along similarlines early in the prologue: nu`n d∆ejcqra; pavnta, kai; nosei` ta; fivltata (16).

54 MURGIA 1987: 56, contra CASALI 1999: 113 n. 22. 55 MACKAIL 1930: 151, endorsed by AUSTIN ad loc.56 An intuition shared by a master of stylistic analysis and – in his youth – a disci-

ple of Freud, Leo Spitzer: ‘To every emotion, that is, to every distancing from ournormal psychological state, corresponds, in the field of expression, a distancing fromnormal linguistic usage; and, conversely, […] a distancing from usual language is asign of an unusual psychological state.’ (SPITZER 1951: I.4). RIFFATERRE 1987, especially371-374, restates the principle in explicitly psychoanalytic terms.

57 Cf. D. P. FOWLER 1997: 25 = 2000: 127-128, quoted in the heading.


we simply cannot write out of morte. No: Dido cannot be fully awareat this stage of the violent impulses she feels against Aeneas, and whichshe will voice much later, just as she cannot (consciously) know thather request for tempus inane connects her with an ominous intertextu-al model.

Awareness of the Euripidean intertext – and a willingness toexplore the meaning of morte to the full – thus add a crucial layer ofsignification to Dido’s words. We therefore have (i) the quasi-tempo-ral, reassuring meaning ‘at [my] death’, which is intended to deceiveAnna and succeeds in doing so because it can be understood to referto a form of future material gratitude towards Aeneas; (ii) a passive-aggressive meaning – ‘my suicide will be Aeneas’ recompense,’ whichAnna is not expected to detect, but – arguably – a well-read Aeneas is;and finally, (iii) a latent aggressive meaning: Dido’s words imply a pos-sible revenge against Aeneas, who will be ‘recompensed with his owndeath’ for his cruel treatment of Dido. This last meaning is also activat-ed by the intertextual memory, but is not supposed to be detected byeither Anna or Aeneas, in case it results in the failure of Dido’s requestto see her lover. We will return to this particular issue 58.

4. The variant reading cumulata is known to Servius (he does notendorse it) 59 and read in the Mediceus, but has not been very popu-lar among modern editors and critics, if nothing else because it canprima facie be dismissed as an haplography 60. It was championed byHeinsius (followed by Burman), who objected to the repetition of

58 Note a comparable plurality of interpretive levels in Medea’s statement at 341,where h|i feuxouvmeqa can mean both ‘by what road we shall go into exile,’ ‘in whatmanner we shall go into exile,’ and ‘in what manner we shall live in exile’: ‘Creon is tounderstand the first and third, the audience may also hear the second’ (MASTRONARDE on341).

59 Servius on 436: ‘male quidam legunt quam mihi cum dederit (id est Aeneas) cumu-lata morte remittam, et volunt intellegi ‘acceptum ab illo beneficium mea morte cumula-bo et sic relinquam’’, an explanation which actually implies cumulatam (SABBADINI 1930,apparatus ad 436). Servius’ explanation may be at the root of LA CERDA’S insertion ofrelinquam in the text instead of remittam.

60 SABBADINI 1900. MURGIA 1987 and CASALI 1999 do not discuss the relative merits ofthe two readings.

75


-am, but offered no comments about its meaning. It was also initial-ly preferred by Henry 61, who takes cumulata morte to signify – metaphorically – 62 ‘a state of misery exceeding death,’ and inter-prets lines 435-436 as follows: ‘I will remit (cease to trouble him withmy love) in accumulated death, i.e. in a condition worse than death’(for the expression he compares – e contrario – St. Th.11.581 morsimperfecta) 63. But Henry subsequently 64 changed his mind, optingfor cumulatam and making no mention of his earlier preference.However, Forbiger’s summary of Henry’s interpretation – ‘morsquasi multiplex’ – probably lies behind Pascoli’s translation of cumu-lata morte 65 as ‘con una morte moltiplicata’ 66 – Dido pictures herselfdying a thousand times over once Aeneas will have left again (as shefears) after granting her the small delay she is desperately seeking 67.Although Henry and Pascoli both insist on the metaphorical value of

61 In 1851, during a visit to Leipzig (J. HENRY 1853: ix-x), Henry shares his interpre-tation with Forbiger, who records it in FORBIGER3 1852, its first public mention. Henrysubsequently discusses it himself in J. HENRY 1853: 61-66 and 1857: 260-261. FORBIGER4

1872-75 mentions both Henry’s original interpretation, and the fact that he has with-drawn it, ‘ut, quid nunc de eo statuat, nesciam.’ CONINGTON3 1876 also discussesHenry’s reading (finding it ‘absolutely impossible’), while CONINGTON-NETTLESHIP4 1884,appearing after the publication of vol. 2 of Aeneidea in 1878 (J. HENRY 1873-92), dropsthe reference altogether.

62 J. HENRY 1853: 64: ‘cumulata’ is added to ‘mors’, not merely to heighten theexpression, but to place it beyond doubt that ‘mors’ is taken, not in its literal, but in itsmetaphorical, sense.’

63 SNIJDER 1972 – apparently unaware of HENRY and Pascoli – accepts cumulata andinterprets morte as ‘a dying away, mortification’ of Dido’s love for Aeneas, who shouldtherefore feel free to linger in Carthage without any fear that their relationship mightresume.

64 J. HENRY 1873-92: 745-748. 65 ‘[L]ezione del Mediceo, la vera’: PASCOLI 1892 ad loc.66 Thus, incidentally, supporting TRAINA’S 1989: 94 intimation that PASCOLI must have

used Forbiger. 67 ‘La frase cumulata morte è volutamente equivoca: deve valere per Anna: – alla

mia morte, quando ne sarà giunta l’ora… –. Ma Didone dentro sé intende: – conmorire non una volta sola, ma più, con una morte moltiplicata –, perché quel tempoche Didone chiede, non farà se non irritare la sua passione e accrescere il dolore perla necessaria partenza, che ogni giorno, pensandola nel futuro, vedrà in atto diavvenire nel presente.’ Anna, on the other hand, will simply interpret cumulata morteas ‘alla mia morte, quando ne sarà giunta l’ora’, taking cumulata in the sense of plena,perfecta, matura (cf. e.g. Virg. Aen. 6.745 perfecto temporis orbe).


mors, and thus refer it exclusively to Dido’s terminal sadness, Pa-scoli’s translation comes tantalisingly close to an interpretation ofcumulata morte – ‘with abundant death’ – which would substantiallyincrease the risk of reading in Dido’s word an unwelcome ‘note ofsavage irony.’ The problem, therefore, is not, as Pease would have it,that cumulata has ‘no very good meaning’ 68, but that such a meaningwould not easily square with the common assumption that Dido’sspeech to Anna is humble and resigned.

Alternatively, cumulata morte could also be an ablative absolute 69.Taking cumulare in the sense of ‘coacervare’ 70 (‘with death piled uphigh,’ ‘with accumulated death’) 71, it would conjure up a strikingvisual image which would further magnify the ambiguity of the con-text, especially if we consider mors here as abstractum pro concreto for‘corpse.’ This use is safely attested in poetry before and after Virgil 72.Although there are no direct Virgilian parallels for it, mors has thematerial sense of ‘gore’ at Aen. 9.348 (multa morte recepit), whereServius paraphrases ‘cum multo cruore’ 73; on the other hand, Virgilpresents several instances of funus = ‘corpse’ 74.

Dido’s cumulata morte would evoke an image of extensive deathand destruction for which Lucretius offers a significant model in hisdescription of the carnage provoked by the plague. At a climactic

68 PEASE on 436. 69 For comparable expressions cf. Liv. 22.2.8 cumulatis sarcinis; Colum. 10.152 cumu-

latis … glebis; Curt. 6.6.29 stipitibus cumulatis (‘ammucchiando tronchi su tronchi’, trad.Gargiulo). Cf. also Luc. 8.729 non pretiosa petit cumulato ture sepulchra.

70 TLL 4.1383.21ff. 71 An interpretation entertained but quickly rejected by both Pease and Mackail. LA

PENNA 1971: 458 also rejects cumulata morte (‘aggiungendovisi la morte’). 72 Cat. 64.362, Prop. 2.13.22, Manil. 4.665. Cf. also Cic. Cluent. 201, Sest. 83, Mil. 86,

and ThLL VIII.1504.41 ff., with HAGENDAHL 1936: 337-338 and HOFMANN, SZANTYR 1965:2.749.

73 HARDIE ad loc. suggests that Virgil’s ‘bold use’ might have been eased by the ana-logy with Gk. fovno", both ‘slaughter’ and ‘gore.’

74 Aen. 2.59, 6.150, 6.510, 9.491. Other instances in Cat., Prop., Ov., Sen. trag., Luc.etc. (ThLL VI.1605.36 ff.). Both mors and funus are used in this sense because theword cadaver was considered unsuitable for poetry: NORDEN on Aen. 6.150, AXELSON

1945: 49, and P. G. FOWLER 1983: 547-548. The influence of Gk. qavnato" must also havebeen at work. Note also that at Eur. Med. 374 nekrov" = ‘corpse’: ‘make three of myenemies into corpses’ (MASTRONARDE ad loc.).

77


point in the final narrative of De rerum natura, the disease spreadswithout hindrance among humans and animals alike, and mutual con-tagion cumulabat funere funus (6.1238) 75. (Note, incidentally, thatDido’s love had been considered a pestis by Juno as early as 4.90). Asimilar image is found in Lucretius 3.71, where he describes the depthsof depravation to which men can sink in their senseless quest for richesand power: sanguine civili rem conflant divitiasque / conduplicant avidi,caedem caede accumulantes 76.

Here funus is used in the concrete sense of ‘corpse,’ as in Liv.26.41.8 (ut aliud super aliud cumularetur funus) but the author of BellumHispaniense will conflate the Lucretian imagery and the concrete use ofmors in a related passage with distinctive poetic overtones 77: non solummorti mortem aggerabant, sed tumulos tumulis exaequabant (5.6). Apartfrom the influence of Lucretius, a possible Ennian model has been sug-gested 78: indeed the image of heaped corpses retains a strong expres-sionistic flavour which would not sound out of place in the Annales,perhaps in a context similar to Acc. trag. 317 R.2 = 138 Dangel nec per-dolescit fligi socios, morte campos contegi, which also depicts a large-scale massacre and features mors in the sense of ‘corpse’ 79.

After Virgil 80, the use of cumulare in contexts describing massslaughter remains productive 81, for instance in Lucan (4.571 strage

75 P. G. FOWLER 1983: 548: ‘cumulare, although common in an extended meaning,strongly suggests a physical heap,’ as explicitly brought out 6.1262-1263 quo magisaestu / confertos ita acervatim mors accumulabat. The image may have influenced Virg.Georg. 3.556-557 iamque catervatim dat stragem [sc. Tisiphone] atque aggerat ipsis / in stab-ulis turpi dilapsa cadavera.

76 On the specific form of this and similar expression cf. LANDGRAF 1888: 162. 77 See the detailed analysis by PASCUCCI 1965: 173-174 and DIOURON 1999: 63. 78 PASCUCCI 1965: 174. On other possible Ennian reminiscences see later nn. 101

and 104. 79 The only pre-I century b. C. attestation of cumulo, however, is from a fragment

by Caecilius Statius (230 R.2) quoted by Cicero (Cic. Cael. 73): nunc meum cor cumulaturira (a stern old man is talking). Cf. Virg. Aen. 4.197 aggerat iras.

80 Where a comparable image can be found at Aen. 11.207: confusaeque ingentemcaedis acervum.

81 The negative use of cumulo to denote the piling up of evils seems also to beborne out by other passages, such as Cic. Verr. 3.85: haec tam parva civitas … cumulataaliis tuis maioribus iniuriis, or Cat. 1.14 nonne etiam incredibili scelere hoc scelus cumulavisti?


cruenta cumulata ratis) and Statius (Theb.10.655 exanimes circumcumulantur acervi), but especially in Silius, where, of course, we con-template the eventual outcome of the rift between Carthaginians andRome: et iam corporibus cumulatus creverat agger (1.418), cumulanturacervo / labentum (14.609: in the context of a plague), and cumulatamstrage virorum (15.406, with cumulatam in the same position as in Aen.4.436) 82. Although a different sense of the verb is implied, Medea’swords at Sen. Med.147 are also worth mentioning: alto cinere cumulabodomum 83. In keeping with the ominous implications of morte, cumula-ta morte would graphically evoke – with the support of impressive andmemorable models – the spectre of extensive slaughter. Mass funeralpyres, corpses accumulated one on top of the other, would be a suit-able form of revenge on the part of the queen who has been horriblywronged, and, shortly later, she will be sorry that she has not inflictedsuch a punishment on Aeneas, his son, his companions and his descen-dants (600-602) 84. The ‘piling up’ of corpses evoked by cumulatamorte is conveyed, in Euripides, by the rhythm and structure of theline where Medea envisages multiple homicides (374-375: trei`" tw`n

82 Silius displays a remarkable interest in the creative potential afforded byoverblown images of masses of bodies. See for instance 8.659-660 iam stragis acervis /deficiunt campi, or (even better) 8.668-669 pons ecce cadentum / corporibus struitur.

83 For other possible connections between the play and Aen. 4 see later, p. 84. 84 Cf. also 612-620. Of course the Trojans’ instinctive feeling as they approached

Carthage’s shores was a deep-seated (and seemingly justified) fear: only divine inter-vention will cause their hosts to set aside their ferocia / … corda (1.302-303), but dangeris always lurking beneath the surface (cf. e.g. 1.539-540), and to Jupiter they remain ‘anenemy people’ (4.235): SCHIESARO 2005: 91-93 cf. NELIS 2001: 71-73. Note the gloomycharacterization (pace AUSTIN’S insistence [on 1.159] on ‘peace and safety’) of the naturalharbour the Trojans find upon their arrival near Carthage, which displays the typicalfeatures and colours of loca horrida, much as it offers, for the time being, a welcomerespite. Mention of the Nymphs, too, is not entirely reassuring, given their later role inDido’s ‘wedding’ (see below p. 108 on the deceptive stillness of water see later n. 122),and the fact that later in the tradition (Ovid and Silius) Anna will have to metamor-phose into a nymph in order to escape Lavinia’s revenge (see later pp. 96-99) thusbringing to a close the story of the Carthaginian sisters. Anna’s new abode as describedby Silius (8.197-198) echoes 1.167-168. Ovid, characteristically, will tease out the nega-tive connotation of the scenery by echoing it as a backdrop for Ino’s suicide at Met.4.525-527. The emphasis on details such as overhanging rocks and great heights inject aBacchic flavour in the description – both Dido and Ino will turn into ‘Bacchae.’

79


ejmw`n ejcqrw`n nekrou;" É qhvsw, patevra te kai; kovrhn povsin t∆ejmovn) 85; inEnnius, a similar effect is reached through the careful balancing ofsymmetrical elements, repetition and assonance 86. In Virgil, by theanaphoras at lines 600-601 and 605-606, where Dido regrets that shehas not killed Aeneas, his son and his companions, and memet superipsa dedissem (606) directly conveys the relevant spatial detail.

5. While cumulata deserves more attention than it normallyattracts, the accusative cumulatam is doubtless the correct reading atline 433, since the image of devastation which cumulata morte conjuresup could not – of course – be explicitly voiced at this stage, or Dido’srequest would collapse. But neither can cumulatam ever fully dispelthe shadow of cumulata and its threatening implications. The phonetictexture of line 436 is striking, and has rightly attracted attention: thenumber of m is so large that Highet 87 suggests it may try to replicateDido’s sobs as she speaks 88. As we hear the line (indeed, to a certainextent, even as we read it) the difference between CUMULATAMORTE andCUMULATAMMORTE effectively disappears 89: word-order and phoneticeffects conspire to make cumulata morte perceptible as a censoredthought inescapably embedded within the accusative cumulatam inagreement with veniam 90. Even as we print cumulatam, the text still

85 Note also that nekrouv" is a predicate: ‘make three of my enemies into corpses’(MASTRONARDE on 374).

86 Cf. LENNARTZ 1994: 213-238, esp. 218-220. 87 HIGHET 1972: 137. 88 Note the ‘perceptible alliteration’ of m (as well as p) in Euripides’ Medea speech

at 340-347, ‘perhaps expressive of [her] earnest tone’ (MASTRONARDE ad loc.). The veryname ‘Medea’ frequently favours alliterative patterns in m, notably in Seneca’s tragedy(TRAINA 1981: 123-129).

89 Cf. ALLEN 1978: 31. The difference in pronunciation between single and doubleconsonants does not prevent puns: AHL 1985: 57. SABBADINI 1900 argues that thesequence caused the ‘errore di copiatura’ cumulata.

90 Romans were fully attuned to the existence of ‘acoustic intratexts’ (SCHORK’S 1996label), as attested – e.g. – by the famous wordplay on cauneas = cave ne eas reported byCic. div. 2.84. See later n. 233. In general, see AHL 1985 for Latin, and the pioneeringwork of SILK 1974: 173-193 on ‘aural interaction’ in Greek (esp. 191-193 on ‘aural sug-gestion’, a category which is relevant here). On Virgil cf. also SIMONETTI ABBOLITO 1995:178-181.


enables us to glimpse the flash of revengeful destruction in Dido’srequest, and an underlying set of emotions of which Dido is as yetunaware.

Dido’s repressed emotions emerge in the shape of an allusion toMedea’s deceitfully meek request and in an image of mass death whichthe reader is allowed to overhear, but which remains safely unavailableto the characters involved. We cannot establish for sure at what levelof consciousness Dido nurtures the feelings of revenge which seepthrough her words. The possibility that 433-436 are a fully-fledgedTrugrede intended to bring Aeneas back and kill him cannot beunequivocally discounted, but neither can her continuing passion.Indeed, her insistent denial at 431-432 (non … oro / nec …) is a trans-parent vehicle for voicing her unsuppressed desire to return to theconiugium antiquum and keep Aeneas at Carthage, and it would behard to dismiss it as a rhetorical ploy. As ever in Virgil’s poem, Dido isa character of staggering emotional intensity and complexity, and pre-cisely because of its depth her love for Aeneas can leave room, ifbetrayed, for violent revenge.

6. Instead of doubling his efforts to leave Carthage, Aeneas, even ifcertus eundi (554), falls asleep, and is visited by Mercury, who admon-ishes him to hasten his preparations because he is surrounded by dan-gers (560-570) 91:

‘nate dea, potes hoc sub casu ducere somnos,nec quae te circum stent deinde pericula cernis,demens, nec zephyros audis spirare secundos?illa dolos dirumque nefas in pectore versat,certa mori, variosque irarum concitat aestus.non fugis hinc praeceps, dum praecipitare potestas? 92 565iam mare turbari trabibus saevasque videbisconlucere faces, iam fervere litora flammis,

91 E. L. HARRISON 1982 and FEENEY 1998 provide valuable readings of this ratherneglected episode.

92 This line is fairly close to Eur. Med. 372; note, incidentally, the syntactical simi-larity between the construction potestas + inf. and the Greek accusative absolute ejxovn+inf.

81


si te his attigerit terris Aurora morantem.heia age, rumpe moras. varium et mutabile semperfemina.’

Mercury’s second appearance in the book has often been criticisedboth for its supposed structural irrelevance and for unjustifiedlyattributing to Dido violent thoughts of revenge, although dolosdirumque nefas could easily be seen as a fitting anticipation of (atleast) 93 her suicide plans 94, and the images of turmoil conveyed bymare turbari trabibus and conlucere faces read as Mercury’s restatementof Dido’s explicit threat at 382-387. Indeed, as the Trojans at seaglimpse from a distance the flames of Dido’s funeral pyre (a tantus …ignis which appears to be engulfing the city’s moenia) they are filledwith dark fears: notum … furens quid femina possit (5.6) 95. And yet

93 Dido’s association with dolus should not surprise. Even if we discount any omi-nous implication, she was after all renowned for the ingenious way in which sheacquired Carthage’s land: in Medieval English the common noun ‘dido’ means, interalia, ‘trick’ (BONO, TESSITORE 1998: 2-3). SCHEID, SVENBRO 1985: 337 argue that Dido’s alter-native name Elissa, to be connected with ejlivssein ‘to deceive’ (as in Eur. Or. 889-894,A.R. 1.463), points to a Greco-Roman origin for the foundation myth of Carthage (butsee n. 24 for an alternative etimology). Note that Gk. Mhvdeia is also traced to theIndoeuropean stem *med, of both mevdw and mhvdomai whose semantic spectrum includes‘to think’, ‘to plot’, and ‘to treat with magic rites’ (cf. Homeric mhvdea ‘thoughts,designs’), and which is in turn connected with the stem *me of mh`ti": see CHANTRAINE

1968: 1959-60 s. v. medeor, and USENER 1948: 160-163. On Roman stereotypes aboutfides Punica see HORSFALL 1973-74, PRANDI 1979, HEXTER 1992: 345-347; on Medea as anarchetype for cunning and magic prowess, and the characterization of Corynth, herhome and Sisyphus’, as the home of sophismata (as in Pind. Ol. 13.52-56) see DETIENNE,VERNANT 1974: 176-179.

94 MOLES 1987: 159: ‘Mercury’s sexist views and warning that Dido is dangerous areaccurate.’

95 Mercury foresees that unless Aeneas departs fortwith the Carthaginians willpursue the Trojan ships (with trabibus = ‘ships’ or ‘oars’), set fire to any of them stillashore, and possibly (PEASE) hurl faces against their ships. The sequence is somewhatconfused, thus anticipating Dido’s irate and incoherent remarks at 592-594: non armaexpedient totaque ex urbe sequentur, / diripientque rates alii navalibus? ite, / ferte citi flam-mas, date tela, impellite remos! It is in any case significant that the destruction of theships by fire is envisaged here as well, after being first mentioned as early as 1.525 (pro-hibe infandos a navibus ignes). It will actually occur in Book 5, where Bacchic, Dido-likeTrojan women do in fact set fire to the fleet (659-666: note that 666 respiciunt atram innimbo volitare favillam echoes 5.3-4 moenia respiciens [sc. Aeneas], quae iam infelicis


‘how far [Dido] was from such a plot may be judged from her unhappygropings in 534f.’ remarks Austin, who then proceeds to state that Mer-cury is simply lying, and Dido ‘has no plots against [Aeneas] to do himpersonal injury’ 96. The situation is arguably more complicated. At astructural level, for instance, E. L. Harrison has persuasively shown thatMercury’s double intervention is modelled on Iliad 24, and the Romangod displays the same degree of independent initiative which charac-terises his Greek counterpart 97. In the Iliad, Hermes first escorts Priamon his mission to Achilles (24.334-338), but once the king falls asleep inthe Greek camp he returns a second time, appears to Priam in his sleep,and warns him that he is surrounded by dangerous enemies (24.679-688). According to Mercury, Aeneas, too, has failed to comprehendfully the violence of Dido’s feelings and their destructive (and self-destructive) potential, and has therefore taken her harsh words at 381-386 as a temporary emotional outburst which should not be taken atface value. The time elapsing between Aeneas’ acceptance of Mercury’sinitial order, and his actual departure after the god’s second interven-tion is crucial to the unfolding of Dido’s tragedy, and is crucially charac-terised as ‘borrowed time’ fraught with anxiety 98. Nor is Mercury’ssecond message ideologically vacuous: as a Jupiter-endorsed purveyorof Logos the god’s appearance marks the stark contrast between thedivine and human views of the events taking place at Carthage 99.

Elissae / conlucent flammis, with conlucent taken up in turn from 4.567 conlucere). TheBacchic connotation of the scene is guaranteed by Ascanius’ words (672-673) en, egovester / Ascanius!, a luckier (if seemingly unnoticed) take on Eur. Ba.1118-1119 ejgwv toi,mh`ter, eijmiv, pai" sevqen É Penqeuv" (the word order is noteworthy).

96 AUSTIN on 563, cf. FEENEY 1998: 122 on Mercury’s characterisation as ‘the archety-pal liar.’ For other negative opinions, and the suggestion that the appearance is notreal, but just a figment of Aeneas’ imagination (as argued e.g. by STEINER 1966: 53), seeE. L. HARRISON 1982: 29-30. PEASE, however, believes ‘the spirit already shown in 4, 384-386, and, still more, the rage soon to burst out […] go far to explain it’ (in a similarvein cf. CONINGTON on 563, and PARATORE on 563). If Mercury is indeed a figment of hisimagination, it would appear that Aeneas has subconsciously grasped the menacelurking in Dido’s words.

97 E. L. HARRISON 1982: 12-15; 32. 98 E. L. HARRISON 1982: 23-24 is very good on the significance of Aeneas’ belated

departure. 99 FEENEY 1998: especially 115-116.

83


Within this Homer-inspired narrative framework, an Euripideanintertext plays a decisive role once more: Mercury’s si te his terrisAurora morantem (568) unequivocally reflects Euripides Medea 352-354 100, ei[ s∆ hJ ∆piou`sa lampa;" o[yetai qeou` É kai; pai`da" ejnto;" th`sdetermovnwn cqonov", É qanhi) possibly via Ennius’ si te secundo lumine hicoffendero moriere 101. This allusion has been noticed by critics since LaCerda, but regarded as an inert verbal borrowing (or, in any event, anon-threatening intertext). Once we appreciate the similarities betweenDido’s and Medea’s requests, however, we must infer that Mercury’salmost verbatim repetition of Creon’s words acquires a much more sub-stantial significance: not only does it valorise in retrospect the presenceof the Euripidean intertext at 433-436, but it also confirms to Aeneasthat Dido is liable to turn into a revengeful Medea dominated by ira.Seneca’s own version of the scene (Medea 297-299) is – again – close tothe Euripidean model and its earlier Latin incarnations 102:

capite supplicium lues,clarum priusquam Phoebus attollat diemnisi cedis Isthmo.

We are now in a position better to understand the harsh tone ofMercury’s words. Aeneas has conspicuously failed to grasp the fullmeaning of Dido’s message, and most importantly the intertextual ref-erence which would have revealed its threatening potential. Aeneas,

100 The parallel was pointed out by DE LA CERDA, followed only by CONINGTON, FOR-BIGER, BUSCAROLI and PEASE. PARATORE points rather to Hom. Il. 18.268-269 (also men-tioned by DE LA CERDA), which may well be in the background to Euripides’ line.

101 This line, quoted by Cic. Rab. Post. 29 as the typical utterance of an imperiousking, was assigned to Ennius’ Medea by Scaliger, followed by Vahlen2 (264-265) andRibbeck (224-225 R.2), but not by JOCELYN 1967: 349, who argues that it could belong toany number of Republican plays. The line, however, does appear to fit unusually wellat this juncture in Ennius’ play (Jocelyn’s reservations are based mostly on issues oftransmission). An Ennian flavour could possibly be perceived in the rare use of trabesin the sense of ‘ships’ at line 566 (Austin ad loc. with bibliography), one of only twosuch occurrences in Virgil (cf. Aen. 3.191), which is first found in Medea exul 208-209Jocelyn: utinam ne in nemore Pelio securibus / caesa accidisset abiegna ad terram trabes(Ann. 616 V.2 = spur. 9 Sk. may be spurious).

102 LA CERDA on 568 referring to Ennius and Seneca: ‘uterque Euripidi haesit.’ Hosi-dius Geta’s centoneary Medea restores Virgil’s line to Creon (102).


clearly, hasn’t read Euripides’ Medea 103, and is therefore as demens(562) as the king, who, in Medea’s own words, had ‘come to such adegree of folly’ (371) that he granted her the delay she was cravingbecause he had failed to comprehend what was at stake 104.

Mercury’s statement that Dido dolos dirumque nefas in pectoreversat (563) 105 is justified by his (unsurprising) superior awareness ofthe complexity of Dido’s emotions and mental processes 106. Even ifthe queen is not yet conscious of her vengeful feelings, her words at435-436 open a window onto an array of conflicting impulses: she isresigned to her fate and yet desires one last chance to revive the rela-tionship with Aeneas; she is ready to threaten suicide to manipulatehim, but she also entertains in earnest the thought of killing herself;she begs for ‘just a bit of time’ yet is also drawn to the example ofEuripides’ irate Medea, even if it will take some more time before shebecomes fully aware of these destructive feelings (600-606).

Metaphors play an important – and underestimated – role in Mer-cury’s speech. Varios … irarum concitat aestus (564) picks up and con-firms the narrator’s almost identical words at 532, saevit amor magnoqueirarum fluctuat aestu 107, and thus portrays Dido in the throes of anemotional storm whose ultimate outcome is unclear. Elsewhere Virgilrelies extensively on aquatic metaphors when describing a character’shesitation between different courses of action 108. Note for instanceAen. 12.486-487: heu quid agat? vario nequiquam fluctuat aestu, /

103 To reverse WILAMOWITZ’ famous dictum about Seneca’s own Medea (1919: 3.162). 104 Cf. also Ennius’ traversa mente at 229 Jocelyn. 105 Cf. Creon’s reaction to Medea’s request in Eur. Med. 316-317: ajll∆ e[sw frenwn É ojr-

rwdiva moi mhv ti bouleuvhi" kakovn.106 Note that dolos is generic, and although we can connect it with the possibility

that Dido might have decided to act upon her revengeful instincts before it was toolate, it can also be seen to refer to the fraus which Dido is plotting behind her sister’sback (4.675). Ovid’s Medea will echo this line at Her. 12.211 viderit ista deus, qui nuncmea pectora versat. Dolus is a also a key element in the characterization of Seneca’sMedea. See e.g. the interaction donis / dolus at Med. 882.

107 Saevio can be referred to winds, seas, storms: OLD s. v. 3a. 108 On the Greek background to metaphors associating mental upset with a stormy

sea (esp. for such verbs as ceimavzw and taravssw), see SANSONE 1975: 71. Cf. also SEGAL

1969: 32 on similes involving the sea used to describe intense passion, and primarilyin reference to Seneca, LOTITO 2001: 15-20 (with further bibliography).

85


diversaeque vocant animum in contraria curae, where Turnus’ deliber-ations are introduced by a variation on Dido’s ‘en, quid ago?’(4.534) 109, and diversae recalls varium. Or again Aen. 8.18-21 110:quae Laomedontius heros / cuncta videns magno curarum fluctuataestu, / atque animum nunc huc celerem nunc dividit illuc / in par-tisque rapit varias perque omnia versat, with varias, again, and ver-sare 111, which Mercury refers to Dido – albeit in a different sense –at 4.563 112. Euripides’ Medea had mined this very same set ofmetaphorical associations 113, which are, to be sure, common enough,but appear to be particularly appropriate in a narrative contextwhere sea and seafaring play a significant role 114.

It is in this specific context that we should read Mercury’s closingstatement. Varium et mutabile semper / femina (569-70) is an easy targetfor criticism because it is usually interpreted as a more or less light-hearted 115 prototype of ‘la donna è mobile’ – and thus would indeedrepresent an unjustified, callous remark vis à vis Dido’s devotion andsuffering 116. I would like to suggest, however, that by expanding uponthe nautical metaphors suggested by concitat 117 and aestus at line 564 118

109 HEYNE, FORBIGER, RIBBECK, PEASE and AUSTIN point to Eur. Med. 502 as a possible(though inevitably not exclusive) model for Dido’s aporia.

110 Cf. DServius’ note on Aen. 8.19: fluctuat aestu: utrumque verbum de mari est.111 For versare used in connection with the sea cf. Aen. 6.362 nunc me fluctus habet

versantque in litore venti.112 Cf. also Aen. 12.665-668, with TRAINA ad loc.113 Seneca’s will follow suit (abundantly): quid, anime, titubas? ora quid lacrimae

rigant / vvaarriiaammque nunc huc ira, nunc illuc amor / diducit? anceps aaeessttuuss incertam rapit; / utsaeva rapidi bella cum vveennttii gerunt, / utrimque fflluuccttuuss mmaarriiaa discordes agunt / dubiumqueffeerrvveett ppeellaagguuss, haut aliter meum / ccoorr fflluuccttuuaattuurr. ira pietatem fugat / iramque pietas – cedepietati, dolor (937-943).

114 MASTRONARDE 2002: 35. 115 FEENEY 1998: 121-122 talks of ‘sprightly misogyny.’ HARDIE 1991a: 14 perceives

instead ‘a deeper and sombre truth behind the rhetoric,’ alerting to the instability ofDido’s role as a female leader.

116 ‘Mercury is lying, as he lied in 563’ (AUSTIN on 569). Cf. LYNE 1989: 48: ‘[Mer-cury’s statement] is flatly untrue and meant to strike us as such.’ Contra, TUPET 1976:263.

117 For concitare of winds stirring up a storm see OLD s. v. 2(a). 118 The adjective varius (variosque irarum concitat aestus), is mirrored at 12.831 in re-

ference to Juno (irarum tantos volvis sub pectore fluctus). On aestus cf. FANTHAM 1975: 8nn. 3-4.


(and prepared by fluctuat aestu 119 at 532) Mercury’s apophthegm ratherreflects a traditional analogy 120 between woman and the sea 121: bothcan turn suddenly and unpredictably from benevolence to rage, and asa consequence their apparently peaceful demean or cannot be trust-ed 122. They are attractive, but dangerous, forces of nature 123. Dido, awoman at the mercy of the elements, can shift from lull to tempest 124 –as lines 566-567, moving on from metaphor to description, attest 125.Propertius 2.9.31-36 offers a pertinent term of comparison – note espe-cially the role played by ira 126:

119 Cf. the use of fluctuo at Aen. 10.680 (with a comparable correspondencebetween inner mental processes and the setting), with S. J. HARRISON ad loc., and of fluc-tus at 12.831, above n. 118. See also BROWN on Lucr. 4.1077.

120 See MURGATROYD 1995, esp. 16-17, who examines some important Hellenisticmodels (above all Meleager, A.P. 5.156). Cf. also Menander, Monostichoi 371, p. 54Jäkel, and Aesop. fab. 168 and 207 Perry.

121 The Virgilian master text of the gender dynamics set in motion by the associa-tion between women and sea is to be found in the storm scene at the beginning of thepoem, where the storm engulfing Aeneas and his fleet is presented as a manifestationof Juno’s doli and irae (1.130), while Neptunus – compared to a pietate gravem ac meri-tis … virum (1.151) – restores order with masculine authority: ille regit dictis animos etpectora mulcet: / sic cunctus pelagi cecidit fragor (1.153-154). Cf. also 7.586-590, where La-tinus – velut pelagi rupes (586) – steadfastly resists his people’s demand to wage war(the masses are a assimilated to women and children for their lack of self-control). Cf.also n. 132 below. I plan to return to this issue elsewhere.

122 Cf. Lucr. 2.555-558: … ut videantur et indicium mortalibus edant, / infidi marisinsidias virisque ddoolluummqquuee / ut vitare velint, neve ullo tempore credant, / ssuubbddoollaa cum ridetplacidi pellacia ponti (with aestus following at 562; note the antropomorphic ridet). Cf.also 5.1004-1005: nec poterat quemquam placidi pellacia ponti / ssuubbddoollaa pellicere in ffrraauuddeemmridentibus undis, and see NOUSSIA 2001: 284-286. Ovid’s Dido expands on the theme atHer. 7.53-56 – hardly innocently: quid, si nescires, insana quid aequora possunt? / expertaetotiens tam male credis aquae? / ut pelago suadente etiam retinacula solvas, / multa tamenlatus tristia pontus habet.

123 See for instance SEGAL’s 1969: 24-33 remarks about the ‘suggestive ambiguities’(32) between destruction and nourishment in Ovidian water landscapes.

124 One may also compare Tib. 1.5.75-76 nescio quid furtivus Amor parat. uterequaeso, / dum licet: in liquida nat tibi linter aqua, if liquida is taken to imply ‘running’, i.e.shifting (as the woman’s favour: cf. line 70 versatur celeri Fors levis orbe rotae).

125 At line 531 rursusque resurgens iconically conveys the notion expressed byingeminant, and may also suggest the toing and froing of storm-waves (for surgere usedof seas and rivers see OLD s. v. 9; for ingemino cf. Georg.1.333 ingeminant austri et den-sissimus imber).

126 Note, also, Propertius’ threatened ‘aggressive’ suicide in the lines immediatelyfollowing (37-39), and the final exclamation sanguis erit vobis maxima palma meus (40).

87


sed vobis facilest verba et componere fraudes:hoc unum didicit femina semper opus.

non sic incerto mutantur flamine Syrtes 127,nec folia hiberno tam tremefacta Noto,

quam cito feminea non constat foedus in ira,sive ea causa gravis sive ea causa levis.

This attitude is expounded at length in Semonides’ iambus ‘demulieribus’, fr. 7.27-42 West 128. The punch-line which rounds off thissection of the poem (42) is much disputed, but if ajlloi`o" is taken tomean ‘changeable’, ‘mutable’ 129, it would represent a very close paral-lel for Mercury’s own sententia 130:

Another [sc. type of woman] he made from the sea; she has two characters.One day she smiles and is happy … But on another day she is unbearableto look at or come near to … Just so the sea often stands without a tremor(ajtremhv"), harmless (ajphvmwn), a great delight to sailors, in the summerseason; but often it raves (maivnetai), tossed about by thundering waves

A similar theme is developed at 2.5.11-13, though there amantes of both sexes arecompared to the mutability of seas and clouds. In Hor. carm. 3.9.22-23 Lydia calls thepoet levior cortice et inprobo / iracundior Hadria.

127 Cf. Sall. Iug.78: nam ubi mare magnum esse et saevire ventis coepit, limum hare-namque et saxa ingentia fluctus trahunt: ita facies locorum cum ventis simul mutatur. Syrtesab tractu nominatae. Cf. also Prop. 3.19.7-8 and 3.24.15-16 (the latter in the context of asustained metaphorical discussion of erotic travails as navigation, cf. FEDELI ad loc.).Aeneas had already had trouble with this treacherous and fast-changing features of thelandscape during the storm at the beginning of the poem (1.111 and 1.146), where anexpression such as furit aestus harenis (1.107) suggests an analogy between the stormyseas and the ira of Juno. Syrtis recurs again at 4.41. On the possible echo of Absyrtusin Syrtis see later n. 216.

128 Horace’s Ode 1.5 (on which see NISBET, HUBBARD) shows the continuing appealof this set of metaphorical associations, as the poem weaves sea and nauticalmetaphors in the description of the love affair between Pyrrha and the puer. While thetone is far removed from the drama of Aeneid 4, and, consequently, the analogy withthe sea only conveys the woman’s emotional unreliability, darker images such asaspera / nigris aequora ventis (6-7) are noteworthy.

129 PELLIZER 1990: 129-130. Note some old translations into Latin quoted ad loc.:‘ingenio … mutabili’ (BUCHANAN); ‘marisque instar indolem semper variam habet’(KOELER); ‘è l’oceano cosa mutabile’ (LEOPARDI).

130 For proverbial expressions about the analogy between the woman’s and thesea’s rage see PELLIZER 1979: 31-32; on the diffusion of Semonides’ text in Hellenisticculture, PELLIZER, TEDESCHI 1990: XLII-XLIII.


(baruktuvpoisi kuvmasin foreomevnh). It is the sea that such a woman mostresembles in her temper; like the ocean, she has a changeful nature (tauvthimavlist∆ e[oike toiauvth gunh;É ojrghvn: fuh;n de; povnto" ajlloivhn e[cei) 131.

The naturalistic metaphor presupposed by Mercury’s statementalso contributes to explain the ‘rather dry and scientific tone’ 132 of hissententia, which aims to express, as befits a god, a general truth 133 (a‘law of nature’) 134 grounded in authority and experience 135.

Criticism of Mercury’s ‘slander’ fails to take into account that hiswords do not come as an isolated sentence, but complete a gradual pro-cess of understanding which has progressed at different speeds forAeneas on the one hand and for the narrator and his readers on theother. Unlike Aeneas, Mercury is aware of Dido’s anguished monologueat 534-552, which betrays a troubled and unpredictable state of mind,

131 Translation LLOYD-JONES 1975: 43 (a different translation in LLOYD-JONES 1975: 73).The whole section 37-42 may however be an interpolation: VERDENIUS 1968: 141, butsee also LORAUX 1978.

132 LYNE 1989: 50, who attributes the tone to Mercury’s contempt for Dido (see latern. 135). But the technical flavour of mutabile finds a parallel in tractabilis, which thenarrator refers to Aeneas at 4.439, yet Anna had used in connection with the weather at4.53 dum non tractabile caelum. Tractabilis, too, is mostly used of physical things (beforeVirgil, already Cic. Att.10.11.3 applies it to persons). Seneca elaborates on Virgil’simages at Ph. 580-582, where the nurse compares Hippolytus’ imperviousness to that ofa rock unmoved by the waves of feminine seduction: ut dura cautes undique iinnttrraaccttaabbiilliiss /resistit undis et lacessentes aquas / longe remittit, / verba sic spernit mea (cf. FANTHAM 1975: 3-4). When Dido and Aeneas meet again in the Underworld, the queen is as resistant asa cautes to Aeneas’ soothing words: illa solo fixos oculos aversa tenebat / nec magis inceptovultum sermone movetur / quam si dura silex aut stet Marpesia cautes (6.469-471).

133 ‘Il predicato al neutro anziché al femminile generalizza’ (SABBADINI on 569). 134 The combination of varius and mutabilis occurs again at Aen. 11.425-426: multa

dies vvaarriiiique labor mmuuttaabbiilliiss aevi / rettulit in melius, in Turnus’ rousing – and solemn –reply to Drances. Though – here – positive in outlook, Turnus’ statement still focuseson the unpredictability of events. Note also Curtius Rufus 4.14.19 breves et mutabilesvices rerum sunt, et fortuna numquam simpliciter indulget.

135 LYNE 1989: 49 detects ‘refinements of unpleasantness’ in Mercury’s use of theneuter forms (cf. CONINGTON on 569, who speaks of ‘contempt’), and the ‘ironic’ use offemina. On the ‘predicative use of the neuter adjective in sententiae’ see MCKEOWN onOv. Am. 1.9.4, with KÜHNER, STEGMANN 1.32 and HOFMANN, SZANTYR 444 (who posits apossible Greek influence), and SABBADINI, above n. 133. The use of femina may also beexplained, however, as a balancing element vis à vis the negative statement of variumet mutabile (perhaps with a concessive colour: ‘even a (respectable) lady is still …’).

89


and barely conceals her resentment both against the Trojans’ periuria(542) and Anna’s reckless advice (548-549) 136. Aeneas’ delay may wellsuggest that he is far from eager to depart, of course, but also surelyimplies that he does not take Dido’s threats (380-87) at face value. Mer-cury begs to differ. Henry’s rebuttal 137 – endorsed by Pease – that Didowas unswervingly loyal to Aeneas, rather misses the point of these lines,since the god is referring here to Dido’s emotional fluctuations andwarns Aeneas that her begging for a reprieve at 435-436 is not necessar-ily her final reaction to his departure. After all, even her welcominggenerosity towards the Trojans had been unexpected, and Venus had tointervene to make sure that Dido would not change her attitude – nequo se numine mutet 138. Now, for a second time, a god is concernedthat the queen’s feelings may undergo a sudden transformation, and, asMartianus Capella explains (5.485), turn ex amore in odium 139.

Part two

7. Dido’s words to Anna at lines 416-436 mark a decisive stage inher emotional journey from love to revenge, and really raise the stakesin the battle of wills with Aeneas. Not only has he decided to leave,thus prompting Dido’s first violent rebuke at 365-387. Now he reiter-ates his decision even in the face of a seemingly minimal request froma woman who appears no longer enraged, but submissive and almosthumble. Even if he is not keen to give Dido a last chance to revive

136 The tone of Dido’s first words to her sister – i, soror (424) – is somewhat sur-prising, cf. DServius on 416: ‘ANNA prope invidiose, quia ipsa nuptias suaserat.’ I hadbeen used only once before in the poem, at 4.381, with evident sarcasm (note PEASE’Sdefensive remarks on 424). Further use in the poem is certainly sarcastic at 9.634, pos-sibly at 7.425-26 (but cf. HORSFALL ad loc.), while 6.546 (twice) is emphatically positive.Cf. BRINK on Hor. Ep. 2.2.76, LEUMANN, HOFMANN, SZANTYR 837 and BESSONE on Ov. Her.12.204, with further bibliography.

137 J. HENRY 1873-92: 2.803. 138 Aen. 1.674, cf. above n. 84. 139 See already HEYNE on 569: ‘h.l. tantum eo pertinet, quod in muliebri animo

summus amor in summam saevitiam et immanitatem verti potest, ut adeo Aeneae desalute cogitandum sit.’ Similarly PASCAL ad loc., criticised by BUSCAROLI ad loc.


their relationship (her primary, if dissimulated, objective) he could atleast grant her a final farewell and receive in exchange the retraction ofher curse 140 and a more peaceful journey toward his final destination.By making Aeneas snub Dido’s request without appeal, Virgil massive-ly orients the readers’ emotional allegiances in her favour and againstAeneas 141. Her subsequent hatred, the final malediction, and its grue-some historical consequences 142, will be presented as the result not somuch of Aeneas’ decision to depart (which can after all be blamed ona Fate he cannot control), but more directly – as Pease remarks – 143 ofhis gratuitous unwillingness to grant her the tempus inane he couldeasily spare without scuppering the main plan.

This is a line of thinking which the text itself encourages. Beforedetailing her request to Aeneas, Dido offers an inverse captatio bene-volentiae as she points out that she is not one of his enemies, specifical-ly not one of the Greek leaders who attacked Troy (425-427), nor hasshe dug up Anchises’ tomb 144. There is no contextual need for Dido torecall these historical truisms 145, even if the alternative strategy of

140 CARTAULT 1926: 324 (cf. n. 139 below); AUSTIN on 436; CASALI 1999: 116. 141 Christian authors will go further, and rewrite the story of a chaste Dido. For a

brief account of readers’ reaction to Dido’s plight see FARRON 1980. 142 CASALI 1999: 118: ‘Aeneas behaves … also in such a way as to provoke the curse

which, in the “fictional world” of the poem, is the cause of the sufferings and futuresorrows of the Trojans in the second half of the Aeneid and of thousands of peoplethrough the centuries of conflict between Rome and Carthage.’ Had Aeneas agreed todelay his departure, Dido would have ‘repaid him with interest’ by turning her curseinto a blessing: CASALI (2004-05): 163-164. This assumes that Dido is to be taken at facevalue when she claims, among other things, that she no longer seeks the coniugiumantiquum, and would be entirely satisfied by some inane tempus.

143 PEASE 1935: 34: ‘After this scene and the failure of her subsequent appeal toAeneas [4.416-449], her unrestrained love turns to an equally intense hatred’; E. L. HAR-RISON 1989: 16: ‘After many futile attempts through Anna to delay Aeneas’ departure,Dido’s mental disintegration begins (405ff.)’ (emphases mine).

144 An act which the Carthaginians had committed (Liv. 26.13.13, on Hannibal), andthe Romans abhorred: it threatened the extinction of the polity (cf. Hor. Ep.16.13-14,with WATSON). The desecration was attributed to Diomedes in an alternative version ofthe story recorded by Varro (Servius on 4.427; D’ANNA 1989: 159-196, and p. 96 below),the same according to which Anna was Aeneas’ lover (cf. 4.421-23). It is significantthat these two references to the ‘Varronian’ plot emerge side by side in this passage.

145 The mention of Aulis is particularly poignant given the analogy between Dido’sand Iphigenia’s situations: see DUBOIS 1976 and SCHIESARO 2005: 95.

91


reminding Aeneas of all the benefits she has actually bestowed on himhas already proved ineffectual (4.317 and 373-376). To a Roman audi-ence, however, Dido’s words would surely have evoked the youngHannibal’s oath to be a relentless enemy of Aeneas’ descendants 146.Silius, ever an astute reader of the Aeneid, adds to what we knowfrom historical sources the telling detail that the oath was sworn onDido’s manes 147, thus making explicit the connection between Dido’sdeath and Hannibal’s hostility which the Aeneid already plots as adirect consequence of Aeneas’ desertion and Dido’s curse at the endof Book 4 (622-629) 148. But while this long-term malediction is onlyreadable sous rature in Dido’s excusatio non petita, its evocation intro-duces an element of dissimulation in the context in which it isuttered. Indeed, Quintilian quotes ‘non ego cum Danais Troianamexscindere gentem / iuravi’ as an instance of aversio 149, a form of dis-simulation through which aut aliud expectasse nos aut maius aliquidtimuisse simulamus (9.2.39) 150. The perception of an immediate threat– that Dido may turn against Aeneas, acting as a Hannibal ante litte-ram just as he will eventually follow in the queen’s steps – is hinderedin this context both by the easily discernible allusion to Hannibal’soath, which points to a distant future, and by the reassuring prece-dents Dido’s words evoke. Upon their arrival, as they beg prohibeinfandos a navibus ignes (1.525), the Trojans emphasize in similarterms that they have no hostile intentions: non nos aut ferro Libycospopulare penatis / venimus, aut raptas ad litora vertere praedas; / nonea vis animo nec tanta superbia victis (1.527-529); and Achaemenides’frank admission that he had been an enemy of the Trojans’ (3.602-603scio me Danais e classibus unum / et bello Iliacos fateor petiisse penatis)paves the way for reconciliation 151.

146 Polyb. 3.11, Liv. 21.1.4, Nep. Han. 2.4, Appian. 6.2.9, Mart. 9.43.9. 147 3.118-119 hanc mentem iuro nostri per numina Martis, / per manes, regina, tuos. 148 A detailed analysis of the oath in TUPET 1980, who rightly insists on its connec-

tion with Dido’s final curse in Aen. 4. 149 According to modern classifications, aversio a materia (LAUSBERG 1969: 244). 150 On Ovid’s take on these lines see CASALI 2004-05: 158-164. 151 The repetition of populare at 4.403 in connection with the Trojans’ preparation

for departure retroactively calls into question the earnestness of their statement. Note


At least since Heinze, critics have been inclined to chart Dido’spsychology as a linear trajectory between distinct, well-defined stages,an organic development from love to hatred: ‘Each new phase in theoutward course of events leads to a new phase in her inner develop-ment; and each of these phases represents as purely as possible oneparticular state of mind, uncontaminated by any other’ 152. Heinze’sassessment clearly reveals his penchant for ‘classic’ dramatic forms andorganic, well-constructed characters 153. After sensing Aeneas’ betrayal,Heinze’s Dido moves in an orderly way from ‘painful surprise’ to‘scornful hatred’, then from ‘humble renunciation’ to ‘despair’, andfinally – after line 604 – from ‘anger and thirst for revenge’ to ‘sublimepeace’154. There is little room, in this scene-by-scene crescendo, for anynon-linear commingling of disparate emotions 155. Heinze’s influentialapproach inevitably favours the systematic occlusion of any threaten-ing overtone from Dido’s words at 416-436 156. At that stage Dido is inher phase of ‘humble renunciation’157, if not outright ‘self-abasement’ 158

and therefore, to repeat once again Mackail’s comment, ‘it would bequite out of place to make [her] end this piteous appeal on a note ofsavage irony’ 159. Outright hostility and feelings of revenge will be allowedonly at a later stage, only once Aeneas’ obdurate denial has forced her into a ‘demented state of delirium’, and she is no longer her normal

also that Achemenides’ spargite me in fluctus vastoque immergite ponto (605-606) antici-pates 4.600-601 non potui abreptum, divellere corpus et undis / spargere?

152 HEINZE 1915: 103. 153 On some aspects of the HEINZE’S cultural background see SCHIESARO 2006. 154 HEINZE 1915: 103. 155 HEINZE 1915: 103: ‘[Virgil] does not describe a gloomy, irregular oscillation of the

emotions: his Dido is not tossed this way and that by the conflict of passions. On thecontrary, the tragedy strides to its conclusion in a clear and controlled fashion.’

156 He had been authoritatively preceded by Christian writers, who, as they recastDido as an exemplum of matronly virtue which comprehensively ignores her plight inAeneid 4, can be seen to inherit and magnify Virgil’s own displacing strategy.

157 HEINZE 1915: 103. 158 PEASE 1935: 35 n. 263. Cf. e.g. CARTAULT 1926: 324, according to whom the

request for a ‘court répit’ ‘c’est une façon de retirer les menaces de vengeance qu’ellea proférées contre Enée’, or PERRET (1978-80) 3.189 ‘Le contenu de cette proposition[433-436] apparaît par contraste avec les menaces de la veille … et les fureurs quirenaîtront le lendemain.’

159 MACKAIL 1930: 151, endorsed by AUSTIN ad loc.

93


self 160. Henry puts it unequivocally: ‘Dido […] was unchangeably anddevotedly attached to Aeneas, whom, if she did not pursue with fireand sword, it was not that his inconstancy did not so deserve, but thather magnanimity disdained, and her still-subsisting passion forbade’ 161.

Awareness that Dido’s words to Aeneas through Anna contain animplicit threat does much to complicate Heinze’s linear analysis 162. Itendows Dido’s character with a more nuanced and tormented psycho-logical profile, one in which conflicting emotions are not neatly dis-tinct from each other but interact in unpredictable fashion 163. Indeed,the Euripidean intertext creates a compelling tension between thedominant motif of Dido’s deepening despair and the slow emergenceof a perceptible – if repressed – sense of danger for Aeneas and theTrojans. Mercury’s much criticised ‘superfluous’ 164 second interven-tion (560-570) acquires a different significance if we assume that fromhis divine (if not omniscient) 165 point of view 166 (and our intertextual

160 HEINZE 1915: 104. According to HEINZE 1915: 110 n. 16, ‘Virgil has portrayed Didoin such a way that the inhumanitas involved in, for example, killing Ascanius and serv-ing his dismembered limbs to his father as a Thyestian meal would not appear natural,however justifiable or not one might consider such an act to be.’ Cf. PEASE 1935: 35‘Only here [sc. 4.600-602, 615-620] appear those characteristics of barbarism and inhu-manity which the Romans were wont, perhaps unjustly, to associate with the peopleof Hannibal’; PARATORE 1947: 11: ‘un vago progetto che Didone agiterà solo al culminedel suo sfogo e già con la coscienza della sua impraticabilità’ (emphases mine).

161 J. HENRY 1873-92: 2.803. 162 It also complicates THOMAS’ 2001: 162 assertion that Dryden’s Dido is ‘more dan-

gerous than the Virgilian text suggests.’ Dryden emphasizes the menacing aspect ofher character, but he takes his cue from the Aeneid. THOMAS’ Dido is unequivocallyunthreatening and sympathetic (cf. 159).

163 As rightly remarked, for instance, by LA PENNA (1984-91) 55: ‘Didone, come laMedea di Euripide e Apollonio, è eroina del dubbio: ci sono ondeggiamenti, sianell’azione sia nei monologhi.’

164 HEINZE 1915: 113 n. 34, albeit ‘as far as the narrative is concerned, but not froma technical point of view.’

165 A concept now usefully problematized by CULLER 2004. 166 Mercury’s first message to Aeneas had already established his access to a supe-

rior source of knowledge, see e.g. the echo of Jupiter’s question at 235 in Mercury’swords at 271. Comprehensiveness and objectivity, however, are not the qualities mor-tals should expect from Mercury (or other gods, for that matter). Here, for instance,the Zephyros … secundos encouragingly mentioned at 562 quickly turn into opposingwinds at 5.2 (fluctusque atros Aquilone secabat), only to return (5.32-33: secundi /


competence) Dido’s words fail to sound entirely reassuring: herentreaties carry too many traces of Medea’s ‘sweet-to-hear’ words 167.Aeneas’ delay is thus suffused with an air of danger, and his early-morning departure becomes narratively motivated – there are dangersin store if he awaits a new dawn at Carthage. The whole tragic colour-ing of the central section of Book 4 is strengthened: there is even moreof a reason for Aeneas to leave swiftly, and no less cause for Dido tofeel angry.

If we accept that Dido’s speech to Anna betrays unconscious feel-ings of revenge, we can gain a fuller image of her psychological com-plexity and development in the second half of the book. Love andhatred, self-humiliation and revenge emerge not so much as polarextremes in a gradual, painful journey of discovery and despair, but asconflicting emotions which repeatedly vie for predominance in Dido’stroubled heart. If we allow that Dido’s psychological evolution is lesslinear than Heinze suggests, we quickly realize that her words to Annaare not isolated in suggesting the coexistence of contradictory, mostly

… Zephyri) once Aeneas (unexpectedly for him) realizes that Sicily ought to be hisnext destination (28 flecte viam velis).

167 Cf. Eur. Med. 316 levgei" ajkou`sai malqavk’ (Creon to Medea), and esp. 776molovnti (sc. Jason) d∆aujtw`i malqakouv", when Medea outlines her revenge. The adj.malqakov", which encodes Medea’s deception, resurfaces in Dido’s mollis aditus (423) –she will exploit Anna’s ‘sweet approach’ in dealing with Aeneas (mollis carries furthercomplex overtones, see n. 174). Ennius makes Medea directly aware of the deceivingnature of her soothing words: nequaquam istuc istac ibit; magna inest certatio. / nam utego illi supplicarem tanta blandiloquentia / ni ob rem (222-227 Jocelyn). Blandiloquentia(not otherwise attested in archaic or classical Latin, as Jocelyn points out) is interest-ing in the light of the words with which Dido (probably, not certainly: M. BARCHIESI

1962: 477-479) addresses Aeneas in Naevius: blande et docte percontat, Aenea quo pacto /Troiam urbem liquerit (fr. 23 Barchiesi). Attempts have been made to build upon thisfragment the image of a Circe-like Naevian Dido, who would wheedle Aeneas intotalking with insinuating charm (blande) and cunning (docte), but see M. BARCHIESI’s(1962) 479-481 caveats. However, in Aen. 1.670-672 nunc Phoenissa tenet Dido blan-disque moratur / vocibus, et vereor quo se Iunonia vertant / hospitia (Venus recruitingCupid) blandis … vocibus focalizes Venus’ worries about the effect Dido is having onher son, and may almost be seen as a gloss on the narrator’s definition of theCarthaginians as bilinguis (1.661): quippe domum timet [sc.Venus] ambiguam Tyriosquebilinguis (see above n. 93 on fides Punica). The Homeric model is Athena’s speech toZeus about Calypso, Od. 1.56-57.

95


repressed feelings. In fact these include not just a desire for revengeagainst Aeneas, but also a degree of resentment vis à vis Anna, both forher role in encouraging Dido to pursue a relationship fraught withdangers (548 his, germana, malis oneras atque obicis hosti) 168, and forher excessive closeness to Aeneas 169. According to an alternative ver-sion of the legend, which Virgil in the main rejects 170, it was Anna, notDido, who attracted Aeneas’ love at Carthage: recollection of this dis-carded story-line 171 allows us to appreciate the full import of Dido’swords at 421-423: solam nam perfidus ille / te colere 172, arcanos etiamtibi credere sensus; / sola viri mollis aditus et tempora 173 noras 174.

While Virgil confines himself to oblique hints at a potentially verydifferent plot-line, the erotic undertones of Anna’s and Aeneas’ friend-ship receive a fuller airing later in the tradition. As they meet again inFasti 3 Aeneas is at pains to point out to Lavinia that his welcoming of

168 As confirmed by Her. 7.191, where Anna is meae male conscia culpae.169 LYNE 1989: 31-32 also sees traces of ‘unfair (and perhaps largely unconscious)

resentment’ in Dido’s characterisation of Sychaeus’ death as a betrayal (4.17, see aboven. 43).

170 Apart from Dido’s words at 421-423, traces of this alternative plot also emergeas the text hints at how similar, indeed almost interchangeable, the two sisters are, see4.8 unanima, with SCHIESARO 2005: 105.

171 About which see D’ANNA 1989: 159-196. See above n. 144. 172 As noticed by Nonius, who includes this passage under colere = diligere (p. 250

M.). LA CERDA ad loc.: ‘An in verbo colere latet aliquid? Hoc verbum pro amare, interdumveteres dixerunt, sicut cultores, pro amatores’, and shortly afterwards, noting the ritualovertones of colere ‘ergo quis sciat, an hic Dido aliquid suspicetur de zelotypia, usaquesit obscuro verbo, et ambiguo, indicans Annae ritualitatem, quae ad se virumpellexerit? praesertim cum addat perfidus, perinde, in me perfidus, quia rivalis meus, ettui cultor.’

173 This line is echoed by Silius’ Anna when she first addresses Aeneas: cui sic verbatrahens largis cum fletibus Anna / incipit et blandas addit pro tempore voces (8.79-80): protempore echoes tempora noras, and blandas fully develops the erotic undertone of mollis(see the following note).

174 Significantly, Virgil shows Aeneas reasoning in terms which are subsequentlyattributed to Anna: sola viri mollis aditus et tempora noras (422-423) echoes temptaturumaditus et quae mollissima fandi / tempora, quis rebus modus (293-299). Mollis (about whichsee also n. 167 above) conveys a sense of insinuating indirectness, bordering inAeneas’ case on cowardice (examples in OLD s. v. 13 (a)), in Anna’s on flirtatiousness(OLD s. v. 16, PICHON 1902 s. v. mollis: as used at Aen. 4.66 est mollis flamma medullas).LA CERDA rightly compares an erotic intertext, Tib. 2.4.19 ad dominam faciles aditus percarmina quaero.


the now derelict exile is the least he can do for the sister of his saviour:‘hanc tibi cur tradam, pia causa, Lavinia coniunx, / est mihi: consumpsinaufragus huius opes. / orta Tyro est, regnum, Lybica possedit in ora. /quam precor ut carae more sororis ames.’ (629-632). But Lavinia, unim-pressed by this display of belated pietas, and clearly well-versed in thefiner points of the tradition, suspects more is at stake than meets theeye (633-636), and reacts by playing Dido furens, rather than Didohospes: non habet exactum quid agat: furialiter odit, / et parat insidias etcupit ulta mori (637-638), a nod to Aen. 4.563 (illa dolos dirumquenefas sub pectore versat) which is followed by the appearance of Dido’sghost warning Anna, as Mercury had Aeneas, that she must leave inhaste what has become an hostile setting. But cupit ulta mori, a trans-parent reworking of Dido’s final curse, further complicates the picture.Aeneas left behind in Carthage, the Romans’ own Thebes 175, a worldof ill-defined boundaries, incestuous tensions, blurred gender identi-ties 176, a household (and a land) ambigua (1.661) both because poten-tially untrustworthy and disturbingly confused 177. He is too similar toDido, in many respects almost a double, or worse, given Rome’sfounding myth, a brother 178; both Dido and her sister are attracted tohim, and he may well be to both; his own son 179 is, at times, a substi-tute object of attraction for the queen 180; Dido herself is a woman per-

175 For Thebes as a locus without ‘a viable system of relations and differences’(including sexual ones), see ZEITLIN 1990; on Carthage and Thebes in Virgil and OvidHARDIE 1990.

176 As well as untrustworthy, Romans are fond of depicting Carthaginians as effemi-nate, their use of tunicae with sleeves representing a favourite polemical target (asexplained by Gell. 6.12): cf. Aen. 9.614-616 (see n. 289), with Enn. Ann. 325 V.2 = 303Sk. and several Plautine references listed by SKUTSCH ad loc.

177 See OLD s. v. ambiguus 5 (a), 6, 7. 178 Further discussion and bibliography in SCHIESARO 2005: 96-97. 179 Initially (1.657-660) a disguise for Aeneas’ own half-brother, Cupid; when Venus

talks to him she refers to Aeneas as frater … Aeneas … tuus (1.667). 180 A trace of this ambiguity may be detected in Silius’ narrative as well. When she

lands in Latium Anna encounters Aeneas and Ascanius together, and while she imme-diately recognizes Aeneas’s well-known face, she falls at his son’s feet (8.73-74). Shethen tells Aeneas that in her despair Dido both embraced Ascanius’ statue and gazes atAeneas’ portrait (8.91-93, with nunc … nunc underlining the parallelism between thetwo gestures). Aeneas replies in kind: iuro caput, Anna, tibique / germanaeque tuae dilec-tum mitis Iuli (8.106-107).

97


forming male roles 181. But can he really leave ‘Carthage’ behind? Nowthat he has fulfilled his destiny, and married Lavinia, we discover thathistory is all too ready to repeat itself. Analogies between Dido’s andAeneas’ lives as exiles are no less striking than those now linking Annato Aeneas 182. This time even their names are uncannily similar 183, andso will be their destiny, since Aeneas also metamorphoses into a numenafter being swallowed by the Numicius (Ov. Met. 14.581-608), the sameriver which has become Anna Perenna’s abode. ‘Feminine’ waters ulti-mately engulf Aeneas as he merges with his double – and Dido’s.

Lavinia, whom Virgil encourages readers to see as Dido’s opposite,turns out to be no less emotional, and just as prepared to plot revengein stealth as she feels threatened. She, too, can hate ‘like a Fury’ (furi-aliter), thus showing that Juno’s power is as momentous in Latium as itwas in Africa (Aeneas himself will eventually succumb), and did notabate after Amata’s suicide 184. In this context, Aeneas’ request thatLavinia treat Anna with sisterly love – precor ut carae more sororis ames(632) – is more ominous than reassuring.

As he stages his own Aeneid Silius adds substantially to the intrica-cies of Anna’s story, although the authenticity of a crucial central sec-tion, known as Additamentum Aldinum (8.144-223), is disputed 185.More explicitly than in Fasti 3, Anna’s reappearance in Punica 8 alertsto the continuing power of repetition and regression. Here Aeneas isliterally ‘seized anew’ – repetitus (8.104) – by his love for Dido, onceagain mediated by her Doppelganger sister, and allows himself a senti-mental walk down memory lane, looking back (8.108 respiciens) at thetragic outcome of his sojourn at Carthage 186. As Aeneas, Anna and

181 The complexities inherent in Dido’s sexual identity and role are highlighted bythe unexpected presence of Caeneus alongside the queen in the Lugentes campi: see G. S. WEST 1980, and n. 369 below.

182 For a Virgilian anticipation of this connection see above nn. 144 and 171. 183 AHL 1985: 309-315. The slope is indeed slippery: ‘Anna’ is in turn easily connect-

ed to ‘Hannibal.’ 184 SPENCE 1999: 88-89 offers perceptive remarks on the parallel role of furor in

Carthage and Rome. 185 A strong defence of authenticity, and further bibl., in BRUGNOLI, SANTINI 1995. 186 Respicere is always a loaded act for Aeneas. It is his failure to look back that pre-

vents him from noticing Creusa’s disappearance (Aen. 2.736-740), and when he does


Dido re-enact their triangle with different roles, Dido’s appearance asMercury-cum-Hector (8.168-183) credits Lavinia with very much thesame aggressive instincts we have been tracing in Virgil’s Dido: ‘his,soror, in tectis longae indulgere quieti, / heu nimium secura, potes? nec,quae tibi fraudes / tendantur, quae circumstent discrimina, cernis? (168-170). Dido, to be sure, describes Lavinia’s train of thought with thesame words Mercury had applied to hers in Aeneid 4: iam tacitas sus-pecta Lavinia fraudes / molitur dirumque nefas sub corde volutat quotesilla dolos dirumque nefas in pectore versat (4.563) 187.

Other sections of Book 4 offer substantial evidence, I believe, ofhostile feelings on Dido’s part, and thus help shape a fuller account ofher psychological complexity. They are (i) the authorial introduction toDido’s words to Anna (412-415); (ii) her dream in the anguished nightafter this last entreaty has been rejected (465-473); and finally, (iii) themagic rites at 504-521. Just as important are two short, poignant pas-sages where Dido voices impossible retrospective wishes, first that ithas not been possible for her to lead her life more ferae (550-552), andthen that she has not hurt Aeneas while it was still possible (600-602).The interpretation of these passages which I have offered elsewhere 188

forms an integral component of the present discussion.

8. Virgil had introduced Dido’s words to Anna at 416-436 with anemphatic authorial statement (412-415):

improbe Amor, quid non mortalia pectora cogis!ire iterum in lacrimas, iterum temptare 189 precandocogitur et supplex animos summittere amori,ne quid inexpertum frustra moritura relinquat.

look it is too late: amissam respexi (741) (for the Orpheus and Eurydice model seeSCHIESARO 2003). So, too, as he leaves Carthage, and looks back when Dido’s pyre isalready burning (5.3 respiciens). Ovid’s Anna will behave in a similar fashion: moeniarespiciens (Fast. 3.566).

187 On Anna in Ovid and Silius see also later, p. 208 (SIFC 2/2008). 188 SCHIESARO 2005.189 Temptare can have erotic connotations:. cf. Tib. 1.2.17 with MALTBY ad loc. and

Prop. 1.3.15 with FEDELI ad loc.

99


which clearly recalls and condenses Apollonius Rhodius, Argonautica4.445-449 190:

scevtli∆ “Erw", mevga phma, mevga stuvgo" ajnqrwvpoisin,ejk sevqen oujlovmenaiv t∆ e[ride" stonacaiv te povnoi te,a[lgea t∆ a[ll∆ ejpi; toi`sin ajpeivrona tetrhvcasi:dusmenevwn ejpi; paisi; koruvsseo, dai`mon, ajerqeiv",oi|o" Medeivhi stugerh;n frhsi;n e[mbale" a[thn.

This striking authorial intervention, with its ‘hymnal flavour’ 191,introduces a pivotal development in both the action of the Argonauti-ca and the character of Medea, the killing of Absyrtus (4.391-481).The unusually strong 192 metanarrative question which follows (450-451)193 enhances the prominence of the apostrophe by stressing theenormity of the crime which Medea has devised and which the narra-tor is now going to describe in detail. Virgil establishes an emphaticlink with his Greek model 194 not only by translating verbatim the firsttwo words of Apollonius’ line as a ‘motto’, but also by faithfully repli-cating the internal echo of scevtlie, which the narrator applies to Erosat 445 after Medea refers it to Jason a few lines earlier, at 376 195. Simi-larly, Dido calls Aeneas improbe at 386 196, and Amor is labelled in thesame way at 412 197.

Medea devises the killing of her brother after she fathoms Jason’sbetrayal. She is furious (4.391), and excogitates a mevgan dovlon (4.421)to stave off her demise (her lover had effectively agreed to return herto the wrath of Aeëtes in exchange for a safe voyage home). Under the

190 Note, however, that the reference to Apollonius, already in LA CERDA and HEYNE,is afterwards mentioned only by FORBIGER, CONINGTON, BUSCAROLI and PEASE.

191 HUNTER 1993: 116 n. 68. 192 LIVREA 1973: 143; PADUANO, FUSILLO 1986: 585; HUNTER 1993: 116. 193 Virgil inverts Apollonius’ sequence, and has the apostrophe (412-415) follow,

rather than precede, the narrator’s metanarrative intervention (408-411). 194 A comparable link may be established – via Aen. 4 – between Apollonius’ line

and Aen. 3.56-57 quid non mortalia pectora cogis, / auri sacra fames: JACOBSON 2005b. 195 The connection between scevtlio" at A.R. 4.376 and Aen. 4.386 does not appear

to be mentioned in commentaries, or in NELIS 2001. 196 As Medea calls Jason at Ov. Her. 12.204. 197 As at Ecl. 8.49 and 50.


combined strength of Eros and Ate (4.445-449) Medea metamorphosesinto a fully tragic character: her insistence on the betrayal of cavri" 198,her worries about the enemies’ scorn (389) 199, the entrapment ofAbsyrtus with devious words and beautiful garments 200, the mention ofAte – these are all details which strongly convey the centrality oftragedy in this part of the epic poem 201. It is exactly at this point thatApollonius’ Medea turns into the protagonist of Euripides’ play.

I suspect that Virgil’s transparent allusion at 4.412 is designed pre-cisely to evoke this momentous development in the story of Medea 202,and to cast an ominous shadow over Dido’s words at 416-436. Asimprobe Amor prepares readers for the threatening implications of hermessage to Aeneas, the allusion can be considered, retrospectively, asthe opening salvo in a sequence of carefully balanced clues whichpunctuate the text for the next two hundred lines, and coherently, ifcovertly, develop the theme of Dido’s intermittent and repressed mur-derous intentions.

Direct as the evocation of Apollonius undoubtedly is 203, Virgil’sstrategic manipulation of his model is evident, too. Improbe Amor is adirect translation of scevtli∆ “Erw" which bypasses Catullus’ sancte puer(64.94), but Virgil abruptly cuts off the text to which he alludes: hewrites out Apollonius’ apotropaic gesture at 4.448-449 204, reduces to aone-liner his insistent description of the sufferings caused by Love (445-447), and shifts the emphasis from the pain that Love can inflict (e[ride",stonacaiv, povnoi, a[lgea) to the actions Love forces to accomplish.

These innovations are all part of a coherent attempt to tame a pas-sage which in its original context heralded the description of the most

198 PADUANO 1972: 217. 199 PADUANO 1972: 222. 200 HUNTER 1993: 60-61. 201 PADUANO 1972: 217, 223-223. Note, however, that Paduano rightly stresses the

fundamental continuity of Medea’s character, who, even as the young woman in loveof Book 3, is never a ‘shy’ Mädchen who suddenly turns into a dangerous sorceress.Cf. PADUANO, FUSILLO 1986: 581. On a different line KENNEY 2001: 267-268.

202 Although, of course, subsequent events differ, as remarked by NELIS 2001: 165. 203 Apollonius’ own opening words also appear to replicate Theognis 1231-1234,

but on the relationship between the two texts see VETTA 1980: 37-39. 204 HUNTER 1993: 116 n. 68.

101


violent of violent crimes. In the Aeneid, improbe Amor is suitablyremoved from the curse Dido utters at 381-387 205, and instead followsthe elaborate description of Aeneas’ preparations to depart and thepain they inflict on Dido; most significantly, improbe Amor introducesDido’s final, desperate appeal to Aeneas through Anna at 416-436.Indeed, the context invites us to recognize that Love is improbus pre-cisely because it can turn a proud 206 woman like Dido into a supplexready to humiliate herself in the vain attempt to retain her unfaithfullover (413-415), not because it can push her to act as Medea did 207. (Itshould be noted, however, that Dido’s self-characterization as a suppli-ant is in itself suspect, for already Euripides’ Medea manages to over-come Creon’s resistance, in the scene we have been referring to, byposing – deceptively – as a suppliant) 208.

While the novel context, and the use of words such as supplex, pre-cando, summittere and frustra (not to mention moritura), strive toemphasize the difference vis à vis Apollonius’ model, they cannot fullyerase its menacing nature, which still lurks in Dido’s request to Aeneas,and alerts the reader that in the queen’s mind love and hate, resigna-tion and revenge continue to coexist: there is more to link Dido’sverbal violence of 381-387 to the desperate appeal of 416-436 thanmeets the eye 209.

Just as the allusion to Euripides’ Medea surfaces twice, at lines 433and 568, so, too, the Apollonian intertext heralded at 412 finds a cru-cial confirmation later in the book. Directly after Mercury spells out toAeneas how dangerous it is to delay his departure (and in so doing

205 Hosidius Geta will push in the opposite direction; it’s Medea herself whoexclaims improbe Amor, quid non mortalia pectora cogis? (12).

206 As conveyed by the plural animi at 4.414 (see AUSTIN ad loc.). 207 By drastically condensing the apostrophe, Virgil also suppresses the sea

metaphors which punctuate Apollonius’ lines (Catullus 64.94-98 has iactastis andfluctibus), and associate the effects of Eros to a storm (HUNTER 1993: 117 n. 72).These metaphors, however, play an important role in the surrounding context, seeabove p. 85. Moreover, as HUNTER 1993: 117 n. 72 rightly notes, at 408-410 Dido islooking out at the sea.

208 The specific language of supplication occurs repeatedly in the scene with Creon(324-351), at 326, 336 and 338. See MASTRONARDE 2002 on 324-351.

209 On the repetition of i at 381 and 424 see above n. 136.


confirms the relevance of the Euripidean intertext to Dido’s messageat 433-436), Dido finally displays an awareness of her murderous feel-ings. By now, of course, it is too late, because Aeneas is gone, andDido can only express her regret at not having acted earlier (600-606):

non potui abreptum divellere 210 corpus et undisspargere? non socios, non ipsum absumere ferroAscanium patriisque epulandum ponere mensis?verum anceps pugnae fuerat fortuna. fuisset:quem metui moritura? faces in castra tulissemimplessemque foros flammis natumque patremquecum genere extinxem, memet super ipsa dedissem.

In this sustained fantasy of counterfactual revenge 211 Dido bringstogether a number of notable paradigms (including Pentheus’ sparag-mos) 212, but lines 600-601 refer unequivocally to the maschalismos 213

of Medea’s brother Absyrtus 214: the repetition abreptum/absumere 215

could perhaps be seen as a phonic reminder of the man’s name 216.Although the most detailed description of the murder is offered in theArgonautica (where Jason strikes the fatal blow) Dido’s words recall adifferent (older) tradition 217, according to which Medea herself is themurderess. This is the version mentioned by Euripides’ Medea (lines

210 Cf. GOSSRAU ad loc.: ‘Baccharum more.’ 211 The late, anonymous Epistula Didonis ad Aeneam (AL 71 Sh.B.= 83 R.) trans-

forms non potui …? into poteram (93, 95): but amor prevented commission of such anefas. I return to Virgil’s passage later, p. 201 (SIFC 2/2008).

212 See later, p. 201 (SIFC 2/2008). 213 A ritual form of slaughter normally practiced by someone who believes to have

been wronged – an assumption more operative for Dido than for Medea at this stage:GLOTZ 1904: 62-64, DELCOURT 1939: 162.

214 ‘Brotherly slaughter’ is evoked, in a quite different, but still intriguing sense, at4.21: sparsos fraterna caede penates. Hosidius Geta’s Medea (9) applies the expression toher own killing of Absyrtus, whose ghost makes an interesting, if all too brief, appea-rance in the play (390-391).

215 Absumere appears to be used frequently of multiple deaths, cf. OCD s. v. 7 (a),(c).

216 A similar effect, as Philip Hardie suggests, may be traced in the early mentionof Syrtis in Book 1 (see above n. 127).

217 BREMMER 1997: 86. Later sources resurrect this version of the myth, see BREMMER

1997: 86 n. 11 (but Callimachus fr. 8 Pfeiffer does not bear BREMMER’S reading: D’ALESSIO

1996: 389 n. 53).

103


167 and 1334), who, however, commits the crime next to the familyhearth before leaving Colchis: by blending the two different strains ofthe tradition Virgil engages at once with both Euripides and Apollo-nius, foregrounds Medea’s responsibility, and places Absyrtus’ death ina marine setting which is appropriate to the narrative context 218.

The final part of Dido’s outburst (604-606) establishes yet anotherlink with the Argonautica 219. Medea, too, had threatened to throw her-self onto a pyre as soon as she had sensed Jason’s betrayal (4.391-393):

w}" favt∆ ajnazeivousa baru;n covlon: i{eto d∆h{ genh`a kataflevxai diav t∆ †e[mpeda† pavnta keavssai,ejn de; pesei`n aujth; malerwi puriv.

The image also reinforces the thematic coherence of this centralpart of the book 220. At 407-409 the verb fervere evoked the Trojans’sedulous preparations which Dido watches in despair (408 cernenti);later, at 566-567, Mercury warns Aeneas that if he lingers any furtherat Carthage he will wake up to see the shores alive with flames(videbis / conlucere faces, iam fervere litora flammis); at 594 Dido, inher frenzy, gives orders that the Trojan ships be set on fire (ferte citiflammas); and finally, at 604-605 she regrets not having burned theships while she could 221.

The Apollonian references at 386-412 and at 600-601 frameDido’s tortuous transition from despair to resignation to hatred byevoking the most gruesome, indeed most tragic section of the Argo-nautica. Within the larger narrative compass framed by these allu-sions, the symmetrical intertextual references to Euripides’ revengefulMedea, at 433-436 and 568, opens a troubling vista onto Dido’s feel-ings and intentions.

218 A.R. 4.391-481; Eur. Medea 167, 1334. Cf. PEASE on 4.600 and MASTRONARDE 2002:47-48.

219 NELIS 2001: 168, with further bibliography. 220 On the imagery of Book 4 (in a different perspective), see NEWTON 1957 and

FERGUSON 1970. In general, OTIS 1963, index s. v. ‘symbol.’ 221 Dido’s impossible desire neatly parallels the Trojans’ deep-seated fear as they

reach Carthage’s shores, when they beg of her prohibe infandos a navibus ignes (1.525).Their subsequent disclaimers (527-529) anticipate Dido’s somewhat puzzling remarksat 4.425-427, see above p. 91.


9. Dido’s unconscious is at the centre of our attention in the sec-tion which immediately follows her speech to Anna. After Aeneas hassteadfastly refused to yield to her entreaties (437-449), distress over-comes the queen, as the first thoughts of suicide begin to crowd hermind (450-451). She performs rituals which yield ominous intimations(452-457), and begins to hallucinate: she seems to hear the voice of herdead husband Sychaeus calling her from the grave (460-461), while anowl emits mournful sounds (462-63). There follows (465-473) adetailed description of Dido’s dream, especially remarkable, to beginwith, because unlike other dreams in the poem it does not display anyostensible prophetic function, but is meant to offer an insight intoDido’s inner feelings. Precisely because of its lack of any practical nar-rative purpose (this is the only ‘symbolic’ dream in the Aeneid) 222 itstands out as a crucial stage in her psychological development.

As a whole, this section of the narrative is closely comparable to anearlier part of the book 223 (lines 54-89) where Virgil describes Dido asshe finally yields to her feelings for Aeneas (54-55: his dictis impensoanimum flammavit amore / spemque dedit dubiae menti solvitquepudorem). After Dido performs a series of rituals whose outcomeremains pointedly unclear (56-67), she is overcome by a sort of Bac-chic frenzy (68-69 uritur infelix Dido totaque vagatur / urbe furens) 224:the narrator compares her to a doe who is chased and eventuallywounded by a pastor (71 agens). Dido experiences a series of hallucina-tions: she cannot sleep (82), constantly sees and hears Aeneas evenwhen he is not around (83 illum absens absentem auditque videtque),and when she sees Ascanius she is reminded of his father (84-85 autgremio Ascanium genitoris imagine capta / detinet, infandum si fallerepossit amorem).

A number of specific textual pointers underlines the connectionbetween the two passages. Note for instance the emphasis on sola at 82and 467; the parallel between relinqui (466) and relictis (82), if the

222 STEINER 1952: 48-51, at 49. 223 For interesting aspects of Virgil’s framing technique in Book 4 cf. E. L. HARRISON

1976: 101-103. 224 On Bacchic elements in various sections of the Aeneid see now BOCCIOLINI

PALAGI: 2007.

105


latter is taken to mean ‘just abandoned by Aeneas’ 225; the chasing ofthe pastor agens the doe (71) and Aeneas pursuing Dido (465 agit;cf. 471 agitatus); the analogy between peragrat (73) and longam … / ireviam (467-468).

The evocation, at this particular juncture, of Dido’s painful fallingin love is especially significant. At one level, the parallelism heightensthe pathos of her present predicament by recalling – just as Aeneas hasirrevocably turned his back on her – the intensity of Dido’s feelings,the high hopes she nurtured (note 55 spem), and the sacrifices she hasmade in order to pursue her love (again 55, solvitque pudorem). Allthis has gone horribly wrong. At the same time, the parallelism invitesa retrospective evaluation of the rites Dido performed at 56-67 inorder to obtain ‘peace’ (56 pacem). This has clearly not happened, andDido’s abandonment of pudor has not gone unpunished. The obviousnegative outcome of the rites described at 450-455 belatedly fills in adetail which had not been mentioned in the earlier scene, namelywhether those rites had actually succeeded. Lines 65-66 refuse to settlethe question of what Dido reads, or believes she can read, in the spi-rantia … exta (64) she inspects, offering instead an ambiguous, inten-tionally double-edged narratorial comment 226. The space thus createdbetween the inconclusiveness of the rites at 56-67 and the explicitnegativity of those performed at 450-455 is the setting of Dido’stragedy. The vates who are unable to help at 65 return to haunt Didowith their ominous prophecies at 464-465. This sinister context inwhich past prophecies and dead husbands return to Dido in themoment of her deepest anguish indirectly raises the issue of her culpa-bility, as readers are reminded that the placatory rites she attemptedfailed to achieve their goal.

Sychaeus plays an unexpectedly large role in this section (457-465):

praeterea fuit in tectis de marmore templumconiugis antiqui, miro quod honore colebat,velleribus niveis et festa fronde revinctum:hinc exaudiri voces et verba vocantis

225 As already suggested by DServius, cf. PEASE ad loc.226 As O’HARA (1993) well shows.


visa viri, nox cum terras obscura teneret,solaque culminibus ferali carmine bubosaepe queri et longas in fletum ducere voces;multaque praeterea vatum praedicta priorumterribili monitu horrificant.

Much as Dido (and the narrator) describe his ‘return’ in ostensiblyaffectionate terms (note for instance 458 colebat and 459 festa), hiscoming back to centre stage can hardly be reassuring. Indeed, as anuncanny revenant – this is his second appearance to Dido since hisdeath – 227 who implicitly reasserts his rights over his former wife,Sychaeus’ voice physically embodies the powerful sense of guilt whichbegins to grip Dido at this stage 228: the text suggests a disturbing con-tiguity between his voces and those of the owl, a bird traditionally asso-ciated with the underworld and the dead, who now emits a feralecarmen. Dido rediscovers the sense of guilt which she experienced, butsilenced, at the onset of her passion for Aeneas. Now this guilt is exter-nalised, and materialised in the ghost of her dead husband, whosememory she has betrayed, and who has come back to reclaim her (460-461 vocantis /… viri) 229. Now the predictions of the vates, longunclear or misinterpreted, resurface as Dido (and the readers) come torealise the glaring absence of any positive omen since the very incep-tion of her love story with Aeneas 230. As Sychaeus calls from the grave,he unexpectedly realizes the adynata which Dido relied upon in herinitial oath (4.24-29):

‘sed mihi vel tellus optem prius ima dehiscatvel pater omnipotens adigat me fulmine ad umbras,pallentis umbras Erebo noctemque profundam,

227 On the first, shortly after his death, see later n. 243. 228 MOLES 1987 adroitly disentangles the issue of Dido’s guilt (cf. also MOLES 1984). 229 Ovid (Her. 7.101-102) significantly emphasises this point: hinc ego me sensi noto

quater ore citari; / ipse sono tenui dixit ‘Elissa, veni!’ (note the force of cito, with its legalimplications).

230 One may wonder whether at line 464 the reading piorum (attested in M, andknown to Servius) doesn’t have more going for it than the near-unanimous consensusof modern editors for priorum (F P Lactantius) would suggest. In support of piorum cf.6.662 pii vates, and MACKAIL; contra, among others, J. HENRY and PEASE.

107


ante, pudor, quam te violo aut tua iura resolvo.ille meos, primus qui me sibi iunxit, amoresabstulit; ille habeat secum servetque sepulcro.’

Lines 457-465 precipitate a poignant comparison between Dido’stwo marriages. The reference to Sychaeus in coniugis antiqui (458)recalls coniugium antiquum at 431, thus establishing a marked con-trast between the lasting bond between Dido and her first husbandand the short-lived coniugium of sorts she enjoyed with Aeneas 231.The verbal coincidence is all the more fascinating because of theambiguity of coniugium antiquum – is Dido referring here to the rela-tionship she used to enjoy with Aeneas in the past 232, before hisbetrayal, or to the ‘old-fashioned marriage’ she is willing to forgo inexchange for a little more time together, a less formalisedrelationship 233? The contrast is further carried through in the decora-tion of the altar: Sychaeus’ is adorned festa fronde (459), while shortlyafterwards Dido will place Aeneas’ portrait on a pyre covered fronde… / funerea (506-507). Finally, the owl singing from the top of thehouse (462 culminibus) a dirge whose lugubrious sound is conveyedby the abundance of u sounds in lines 459-462 234, invites comparisonwith the ululations of the Nymphs from above the cave where Dido

231 Silius glosses coniugis antiqui with curas … priores at 8.146-147 (part of the Addi-tamentum Aldinum, see above p. 98): me quoque fors dulci quondam vir notus amore /expectat curas cupiens aequare priores (aequare hints phonically at antiqui).

232 When Dido meets Sychaeus in the Underworld, he is once again her coniunx …pristinus (6.473): at least in death, she has been able to undo history.

233 In such a densely textured set of lines it is not out of place to notice onemore instance of polysemy: quod prodidit oro intriguingly conceals – yet voices –the ‘unconscious’ (and metrically impossible – not a problem for wordplay [AHL1985: 56]) quod prodidi TORO (an anticipation of 4.496-497 lectumque iugalem / quoperii; for torus as an instrument of betrayal cf. Prop. 3.20.25 pollueritque novo sacramarita toro). For a comparable example of ‘acoustic intratext’ at Aen. 7.122-123 (hicdomus, haec patria est, genitor mihI TALIA namque / (nunc repeto) Anchises fatorumarcana reliquit) and 4.408 (quis tibi tum, Dido, cernentI TALIA sensus) see SCHORK

1996. At 4.408 Servius tantalisingly remarks: totum hic magna prosphonesi dictum est:plus enim est in re quam in verbis: quamvis enim totum dictum non sit, tamen et cogi-tatur et capitur ab auditore.

234 Also noticeable in Ennius Ann. 45-46 V.2 = 43-44 Sk.: voce, videtur, verbis.


and Aeneas consummated their union (168 summo vertice) 235. Thewhole section, in other words, is shot through with the typical fea-tures of a well-established theme, the wedding-as-funeral 236, whichwill also play an important role in Dido’s dream 237.

ALESSANDRO SCHIESARO

[continues]

235 Cf. Ov. Met. 6.430-432 (the wedding of Procne and Tereus): Eumenides tenuerefaces de funere raptas, / Eumenides stravere torum, tectoque profanus / incubuit bubo thala-mique in culmine sedit.

236 SEAFORD 1987. 237 See later, p. 197 (SIFC 2/2008).

109

furthest voices in virgil’s dido i · furthest voices in virgil’s dido i ‘we have to stop...

Documents